It '.'.¢' 65:4. ' . W‘ I. ." ' If: "., , 7' ~1J'.‘ 'I‘W 'u " a «a. ‘ :53: ”.34; ‘l j':‘~;n; ”'21“. t. .'t g I .,}"\I-"‘\\“ t ’19, r ,::l" 42:11 .‘I'JN"J 3th “3-2." £45313 ". \' a" ,u ”nil.- ’ji'”..'.;u" we «.01. ll - 3:. 4': -v1 ‘P” .1 'O £0 b.- " V. 15"“: -. 5‘. L. II “'1 ,I:‘ 'a I :;‘ '_.',, “.(."‘A N ‘ l:" '4‘} I $‘;.W.D “8453': '7'" _'.',:;;‘.;S:€§.5;£_ n . Hc‘ :- v33“ km a? 2. av; "s': av!" ,4qu Sv-‘Ji 12 Il‘ .‘ ""'|a':'.‘.$ .,.. .,: 9:: A» ‘sfi "vi" ' (‘3'. ." ‘."" '33:! \ 'v "‘ ""t :.~:~ ‘ - '.V‘~"‘;‘$"7.i‘ I" a I! ;’.'J,'!' 3%ij ' ‘ '7‘ ‘1 ‘,. ‘. . ‘I" - fi, i 1'\\_\':\‘.;,‘,"v a ., -, ‘ ""Q ' “ ‘ I I ; l- as“ 'v. ‘0 , ‘I "g? 1'). W ”I '¢‘{:“ +.3L rV'r . ‘ v ‘n < "I' It (a. ‘ .. _I.-~' ‘ ‘ 'V'IA H. "- o-,‘(' , H ‘ n ;. ‘IH‘. “‘ .'P"la Jam”, . I" ”S: . ‘ x .1’ ins 2133:. m 3: *AW‘ ’- .. "u!" r H‘. 5:. I’I‘Ilg..'%g. ”fink 3:5,} . w I ‘35:}. .53 , _ .,-.\. . I.“ .-.~ |\..‘(\..ll .\ ' ju-t‘ H ' \‘ II ‘;-Q- t": \4 .\ ‘ an «‘0‘ ‘ «vhf: Iggy..." , -“1., ,. ‘x'rd “dup' v-vV. ’H'i " 0"; “J5. TQU "44ng 5:11. ‘ «2459!. Date '11- f‘f (I; s ‘; Y z, 1 i A Michigan State “5‘ University u gum; lllzllliflljllll Ln; my, n14} 1,11 ”mum; This is to certify that the thesis entitled COMPARATIVE ADVERTISING: A DETERMINATION OF EFFECTS presented by ROBE RT THOMAS STACK has been accepted towards fulfillment of the requirements for Ph.D. degree in Marketing 8 Trans. Adm. MW Major pr fessor 11-8-78 0-7639 OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. 1191/1 'Z/ I f f 1' ' 0"} .13.. ‘l *L. . ‘r ’ > a. : ""‘ r “ C ' ' '1 “r‘ 'e',‘ j mm 1 19s». \ I,’ "t ””4 C 3 {if C) Copyright by Robert Thomas Stack 1978 COMPARATIVE ADVERTISING: A DETERMINATION OF EFFECTS By Robert Thomas Stack A DISSERTATION Submitted to Michigan State University in partial fquiTTment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation Administration 1978 ABSTRACT COMPARATIVE ADVERTISING: A DETERMINATION OF EFFECTS By Robert Thomas Stack Until recently, advertising which contained explicit verbal and/or visual identification of competitors was dis- couraged through various industry and media codes and by a common apprehension that naming competitors or mentioning opposing claims served only to promote the competition and to weaken one's own claims. With increased usage and greater explicitness in naming competitors and making clear, by means of comparison, the specific differences between brands, has come a debate over the legal and ethical aspects of comparative advertising; however, little attention has been devoted to systematically examining the effects of comparative advertising. This study attempts to answer some questions of comparative advertising effectiveness confront- ing business strategists and public policy makers. It attempts to determine whether there are any significant differential effects achieved by a comparative versus a more traditional approach to advertising. Hypotheses were developed for six research dimensions: interest level, believability, informational value, sponsor identification, composition of the evoked set, and statement of choice criteria. A conceptual framework, which recog- nizes the interaction of comparative and message sidedness Robert Thomas Stack components, was used to examine variations in advertising format for each research dimension. Four experimental television advertisements, for an existing brand of digital watch, were prepared to represent four treatment conditions: exposure to a one-sided non- comparative advertisement, exposure to a one-sided compara- tive advertisement, exposure to a two-sided non-comparative advertisement, or exposure to a two-sided comparative advertisement. The only element of difference between the four experimental advertisements was the format in which the brand claims were presented; two-sided advertisements dis- claimed superiority on some product features and comparative advertisements identified competitor brands. To minimize the possibility of subjects having a predominantly evaluative response set, such as has been the case where nothing but the advertisement was presented for their evaluation, the experimental advertisements were edited into a syndicated television program and a post test - only methodology was employed. After viewing the televi- sion program, the 125 subjects completed a questionnaire consisting of 18 multiple choice and Likert-type questions. The data generated by the experiment was subjected to the Kolmogorov-Smirnov two-sample test. In analyzing the data it was found that comparative advertisements consistently received significantly better interest level, believability, and informational value ratings than their non-comparative counterparts, with one- Robert Thomas Stack sided treatment showing the most conspicuous differences. No significant differences were found, however, between the interest level, believability, and informational value rat- ings of one-sided and two-sided comparative advertisements. In an apparent refutation of the idea held by many advertising practitioners, that viewers may become confused by comparative advertisements and attribute the sponsor's superiority claims to one of the named brands, the ability of the viewer to correctly identify the sponsor of the advertisement was not significantly affected by variations in message sidedness or comparativeness. Comparative advertisements did significantly influence the vieWers' identification of major competing brands, while the message sidedness and brand ownership components exhibit- ed little or no effect on the incidence of named brands in the composition of the evoked sets of viewers. Similarly, choice criteria for digital watches wen; as. significantly altered by changes in advertising format or brand ownership. Making specific brand attribute claims, via comparison with named and recognizably presented competitors, proved to be a viable technique for improving the interest level, believability, and informational value ratings of a particu- lar promotional message. Evaluating comparative advertising effectiveness is, however, a complex issue and much remains to be learned abou the conditions of comparative adverti- sing effectiveness; a number of future research suggestions are included in the study. To Kay For her patience and understanding ii ACKNOWLEDGEMENTS The road from developing a research proposal to writing a dissertation is lined with people who are willing and eager to help. Most importantly, perhaps, are those who offer encouragement and direction to begin in the first place. I am grateful, therefore, to Dr. Gilbert Harrell who was always available for advice, guidance and encourage- ment, and to Dr. Gordon Miracle for his interest in the work presented here and his many valuable comments. A special thanks goes to Dr. Donald Taylor for reading the manuscript and providing helpful suggestions. Similarly, the technical assistance of Gary McQuaig and his staff at ITV was deeply appreciated and is gratefully acknowledged. On a more personal side, I would like to acknowledge the encouragement and guidance provided by Dr. w. J. E. Crissy; truly a great educator, he reamins ever present in our work and our hearts. Finally, I would like to thank my family for their love patience, understanding, and support. To my wife, Kay, and my parents, a simple thank you does not seem enough. TABLE OF LIST OF TABLES, LIST OF FIGURES INTRODUCTION, Definition Frame of reference Nature of the problem, Research questions THEORETICAL FRAMEWORK , . , Introductory comment , Information processing Message sidedness, , , , Comparative dimensions , THE RESEARCH DESIGN , , Introductory comment , Hypotheses . . . . . . . Methodology, The questionnaire, Data analysis, Concluding comment , RESULTS OF THE RESEARCH , Introductory comment , Interest level iv CONTENTS Page . . . . . v1 . . . . vii . . . . . . 1 . . . . . . l . . . . . . . . . . 2 . . . . 10 . . . . . . 11 . . . . . . . . 12 . . . . . 12 . . . . . 12 . . . . . . . 21 . . . . . . . 28 . . . . . . 38 . . . . . . . 38 . . . . . . . . 38 4O 44 . 45 46 47 47 47 Believability. Informational value. Sponsor identification Evoked set Choice criteria. Selected sample characteristics. SUMMARY AND CONCLUSIONS . Introductory comment Interest level Believability. Informational value. Sponsor identification Evoked set Choice criteria. Recommendations for future research. APPENDIX A (AUDIO/VISUAL SCRIPTS) APPENDIX B (PROGRAM CHARACTER AND SEQUENCE) APPENDIX C (QUESTIONNAIRE). BIBLIOGRAPHY. Page 51 54 58 61 66 72 75 75 75 76 78 79 80 81 82 85 92 94 101 LIST OF TABLES TABLE Page (1) Interest Level and the Comparative Dimension. . 48 (2) Interest Level and the Sidedness Dimension. . . 50 (3) Believability and the Comparative Dimension . . 52 (4) Believability and the Sidedness Dimension . . . 53 (5) Informational Value and the Comparative Dimension . . . . . . . . . . . . . . . . . . 55 (6) Informational Value and the Sidedness Dimension . . . . . . . . . . . . . . . . . . 57 (7) Sponsor Identification and the Comparative Dimension . . . . . . . . . . . . . . . . . . 59 (8) Sponsor Identification and the Sidedness Dimension . . . . . . . . . . . . . . . . . . 6O (9) Evoked Set and the Comparative Dimension. . . . 62 (IO) Evoked Set and the Sidedness Dimension. . . . . 64 (11) Evoked Set and Brand Ownership. . . . . . . . . 65 (12) Choice Criteria and the Comparative Dimension . 67 (13) Choice Criteria and the Sidedness Dimension . . 69 (14) Choice Criteria and Brand Ownership . . . . . . 7O (15) Directionality and Significance Level Results . . . . . . . . . . . . . . . . . . . 71 vi LIST OF FIGURES FIGURE (1) 4A's Comparative Advertising Guidelines. (2) Framework for Conceptualizing Argumentative Structures . . . . (3) Experimental Design. . . . . . . (A1) One-Sided Non-Comparative. (A2) One-Sided Comparative. (A3) Two-Sided Non-Comparative. (A4) Two-Sided Comparative. . . . . . . . (B1) Program Character and Sequence . vii Page 30 43 86 88 89 91 93 CHAPTER I INTRODUCTION In general order of importance, any buying situation presents three questions to the would-be buyer: (1) What is the product and what will it do for me? (2) What does it cost? (3) How does it compare with other products or brands with the same or alternative properties and/or functions? To answer these questions, the consumer accumulates product information from a variety of sources, with much of this information arriving in the form of advertising messages. A relatively new form of advertising message is comparative advertising; it attempts to pose alternative brand choices and buying criteria in specific terms, thereby providing information on the question of how one product or brand '_.__—- compares with other products or brands. Definition The term comparative advertising is not uniquely defined; however, for purposes of this study, comparative advertising is defined as advertising that: (1) Compares two or more specifically named or recognizably presented brands of the same generic product or service class, and 1 (2) Makes such a comparison in terms of one or more spec1f1c product or service attributes. 1 This definition of comparative advertising thus excludes advertising which claims to be "better" or "best," but without making clear to the audience the brands against which the comparisons are made or the attributes on which it is claimed to be superior. Frame of Reference Comparative advertising is not new; as far back as the early eighteenth century printed advertisements named competitors.2 Until recently, however, advertising which contained explicit verbal and/or visual identification of competitors was discouraged through various industry and media codes and by a common apprehension that naming compe- titors or mentioning opposing claims served only to promote the competition, to give them free time or space, and to weaken ones own claim.3 One of the earliest and most popular forerunners to present-day comparative advertisements was the "Brand X" 1William L. Wilkie and Paul W. Farris, “Comparison Advertising: Problems and Potential," Journal of Marketing, (October 1975), p. 7. 2Stanley M. Ulanoff, Comparative Advertising: An Historical Retrospective, (Working paper), (Cambridge, Mass: Marketing Seience Institute, February, 1975), p. 8. 3"Naming Competitors in Ads: Forthright, Fair, Foolish?" (Commentary), Printers' Ink, (January 28, 1966), p. 34; Ulanoff, p. 4; Thomas E. Barry and Roger L. Tremblay, "Com- parison Advertising: Perspectives and Issues," Journal of Advertising, (November 1975), p. 18. commercial which created a hypothetical competitor, which may or may not have had the characteristics of an actual competing product, against which the manufacturer would compare his product. Other early comparative-type commer- cials actually mentioned the competition, but used a "beep" sound that electronically obliterated the announcer's mention of the competing brand; some even had the timing of the "beep" intentionally set off so that the listener distinctly heard the name of the other competing brand.4 More recent attempts to get around the taboo of naming the competition were commercials that depicted the unlabeled, but familiar, package or container of the competitor or used a phrase closely associated with the competitor. Probably the best remembered of these were the Avis-Hertz commercials in which Avis compared themselves to the thinly disguised "Number 1"; similarly, Volkswagen was often referred to "the leading foreign import," and Listerine has frequently been characterized as "the medicine breath mouthwash." Retro- spectively, the early attempts at comparative advertising relied on an indirect kind of competitor identification, comparison by inference, not explicit verbal and/or visual competitor identification. Ulanoff suggests that comparative advertising began in earnest a few years ago, when, in 1971, the Federal Trade Commission's Director of Consumer Protection, Robert Pitofsky, 4Ulanoff, pp. 12-13. asked the ABC and CBS networks to join NBC in accepting comparative advertising and strongly encouraged advertisers, using commercials with a comparative format, to name competing brands.5 Presumably, the commission had a double motivation - to encourage the provision of more factual product information to the consumer and simultaneously to discourage deception by eliminating comparison by innuendo (the contention being that "Brand X" and "another leading brand” formats were too ambiguous since they could create the impression that the advertised brand is better than all competitive brands, when, in fact, it is better than only one of them, or that the unknown brand is the industry leader, whereas, it may be a lesser brand).6 Although court cases established that comparative advertising was legally permissable so long as it did not contain misrepresentations or create a likelihood that purchasers would be confused about the source of sponsorship of the advertiser's product,7 full acceptance and legitimi- zation of comparative advertising did not occur until 1974, when the National Association of Broadcasters and the 51bid., p. 17. 6"Open Way for Comparative Ads, Canada Meet Told," Advertising Age, (May 12, 1975), p. 6; Anthony C. Chevins, "A Case for Comparative Advertising," Journal of Advertising, (Spring 1975), p. 33; Barry and Tremblay, p. 16. 7Sidney A. Diamond, "Court Okays Use of Competitor's Trademark in Ads," Advertising Age (November 16, 1968), p. 66; "Channel Wins OnTy PartiaTTVictory in Unfair Competi- tion Case," Advertising_Age, (November 5, 1973), p. 54. 5 influential American Association of Advertising Agencies, in a complete turn-around from its traditional policy of discouraging the use of comparative advertisements by member agencies, issued revised ethical codes recognizing the phenomenon as an accepted industry practice, that, when used truthfully and fairly, can provide the consumer with needed and useful information.8 This regulatory posture has been supported by various consumer organizations which liken the role of comparative advertising to that of Consumer Reports (which compares specific brands along Specific product attributes). Not satisfied that comparative advertising techniques were being given their fullest opportunities, the Federal Trade Commission authorized its staff to determine whether any existing private codes or self-regulatory review procedures were seriously restricting informative comparative 9 In addition to the major regulatory advertising. organizations - the TV networks, the National Association of Broadcasters and the National Advertising Division/ 8Policy Statement and Guidelines for Comparative Advertising, (New York: American Association of Advertising Agencies, Bulletin Number 3151, April 3, 1974); see Figure 1 for Pattern-Setting 4A's Guidelines. 9Stanley Cohen, "Widespread FTC Probe Will Seek Codes that)Hinder Comparative Ads," Advertising Age, (February 23, 1976 , p. 1. 10. The intent and connotation of the ad should be to inform and never to discredit or unfairly attack competitors, competing products or services. When a competitive product is named, it should be one that exists in the marketplace as significant competition. The competition should be fairly and properly identified but never in a manner or tone of voice that degrades the competitive product or service. The advertising should compare related or similar properties or ingredients of the product, dimension to dimension, feature to feature. The identification should be for honest comparison purposes and not simply to upgrade by association. If a competitive test is conducted it should be done by an objective testing source, preferably an independent one, so that there will be no doubt as to the veracity of the test. In all cases the test should be supportive of all claims made in the advertising that are based on the test. The advertising should never use partial results or stress insignificant differences to cause the consumer to draw an improper conclusion. The property being compared should be significant in terms of value or usefulness of the product to the consumer. Comparatives delivered through the use of testimonials should not imply that the testimonial is more than one individual's thoughts unless that individual represents a sample of the majority viewpoints. FIGURE 1 4A's Comparative Advertising Guidelines National Advertising Review Board - the sweep covered lesser units such as the Distilled Spirits Council and its code.10 This investigation into comparative advertising self—regulation resulted in interim recommendations, including a staff recommendation to establish guides or rules for advertising which mentions competitors by name.11 Increased usage and acceptance of comparative advertising has not, however,been unquestioning. About the same time the Federal Trade Commission was looking into restraints on comparative advertising imposed by industry organizations, the National Advertising Review Board was developing a "white paper" on comparative advertising. The panel concluded that comparative advertising can be benefi- cial to consumers when comparisons are of significant difference and truthfully stated, however, comparative advertising can also increase the amount of misleading advertising and create confusion if used imprOperly.12 Thus 10The codes use such devices as prohibiting direct comparisons or the naming of competitive brands, discourag- ing the use of comparisons or severe restrictions on dis- paragement, establishing more stringent disclosure substantiation requirements for comparative advertisements, and limiting the use of claims of parity in advertisements. 11"Comparison Ads Will Get Little FTC Interference," Advertising Age, (June 5, 1976), 8; Stanley E. Cohen, "FTC Says Guides Near on Ads that Name Names," Advertising Age, (June 28, 1976), 3+. 12Giges, Nancy. "NARB Paper Ducks Conslusion on Comparative Ads," Advertising Age, (August 22, 1977), 62; "Bickering, Fear of FTC Rendered Report Toothless," AdvertisingAge,(August 22, 1977), 62; "Study Cites Value of Comparative Ads But Warns Effect Hinges on Honesty," Broadcasting, (August 29, 1977), 52. critics argue that the undisciplined or uncontrolled use of comparisons in advertising often confuses, rather than clarifies, a buying situation by fostering an environment of counter-claims and contradictions as named competitors in each advertisement attempt to retaliate by selecting (often totally insignificant) product attributes on which their product can be compared favorably with other competing brands;13 that comparisons in advertising, by their very nature, can distort the facts and, by implication, convey 14 In the information that misrepresents the truth. celebrated and landmark Schick Flexamatic Shaver Case, wherein the commercial purported to demonstrate the shaving superiority of the Flexamatic over Norelco, Sunbeam, and Remington shavers via a shaving test conducted by the Nationwide Consumer Testing Institute, the National Advertising Review Board concluded that the Schick adver- tisement was "false in some details and misleading in its 13"Naming Competitors in Ads: Forthright, Fair, Foolish?" p. 34; Sid Bernstein, "Comparison Ads Get Out of Hand," Advertising_Age, (March 29, 1976), p. 16; John M. Trytten, "It's as Easy as Pie: Nothing Can Compare with a Bad Comparative Ad," Sales & Marketing Management, (July 12, 1976), 61-62; Howard GT Sawyer,“Boo to the 'One Book Buy,‘ Ads for Competition, and 'Right?'" Industrial Marketing, (November 1976), 55; Nancy Giges, "Caca-Cola - ReTuctant Entrant into Comparative Ad Warfare," Advertising Age, (January 3, 1977), 2+. 14"Competitors Hail NARB for Schick Shaver Ruling," Advertising Age, (January 7, 1974), p. 1. overall implications."15 Essentially Schick's advertising claims were judged misleading because they implied that the Flexamatic was superior in all cases (where the advertising concerns a product with multiple qualities and character- istics it is possible to establish superiority of one characteristic in such a way that the consumer is led to conclude overall superiority). The review board stated that although the testing technique was basically sound, the Schick advertisements were false and misleading because the differences in the measurements were so minute that they might not be of Significance to the consumer (though the advertisements implied they were), because the advertisements were not sufficiently clear as to the shavers with which the Flexamatic was compared and as to the nature and extent of the tests.16 Maybe the heart of the comparative advertising controversy lies with two key words: compare and disparage. Compare means to examine the character or qualities of something, especially in order to discover resemblances or differences. Disparage means to lower something in rank 17 or reputation, to degrade. Advertisers must not only be 1‘5"NARB Rules Against Schick Shaver Ads, Cites 'Implications'," Advertising Age, (December 31, 1973), p. 1. 15Ibid. 17Webster's New Collegiate Dictionary, (Springfield, Massachusetts: G. & C. Merriam Company, 1973), pp. 229, 329. 10 concerned about the moral question of discrediting a competitor's name, but of disparaging a competitor's goods in such a way that trade libel ensues (Norelco, for example, initiated legal action in the Schick Flexamatic Case because it said it felt "defamed"). Many people in advertising feel that in using comparisons in advertising, advertisers open a pandora's box in the areas of libel, Slander, unfair competition, false advertising and disparagement.18 Nature of the Problem \ Few advertising quandries have so consumed minds from Madison Avenue to Washington as has the debate over compara- tive advertising. It is a complex issue, not subject to generalization or traditional restraint. Since the early years of comparative advertising, experts have differed in their judgments as to its effectiveness. The old axiom - that naming competitors in advertisements doesn't pay - has been replaced by an uneasy suspicion that it is paying. Studies by Shimp (1974), Gallup and Robinson (1975), and Brown and Jackson (1977) indicate that at least seven percent of all television commercials are comparative.19 18Barry and Tremblay, p. 18; Chevins, p. 32. 19Terence A. Shimp, "Comparative Advertising in Nation- al Television Commercials: A Content Analysis," (Paper Presented to the American Marketing Association Educator's Conference, Rochester, New York, August 1975); "Comparative Ads on Rise, Researcher Says," Advertising Age, (March 1, 1975), p. 67; Steven W. Brown and Donald W. Jackson, "Com- parative Television Advertising: Examining Its Nature and Frequency," Journal of Advertising, (November 1977), 17. 11 However, despite increased usage and greater explicitness in naming competitors and making clear, by means of comparison, the specific differences between brands, the_debate over comparative advertising has focused on its legal and ethical aspects, and littleattention has been devoted to systematically examining the effects of comparative adver- tising. As a possibly powerfUl communication technique, it requires careful study and investigation; one cannot assume that just by declaring comparative advertising permissable it becomes desirable or effective. Research Questions This study attempts to answer several questions of comparative advertising effectiveness confronting business strategists and public policy makers. It attempts to determine whether there are any significant differential effects achieved by a comparative versus a more traditional format. More precisely, the following questions establish the boundaries of this study: .How does advertising format affect viewers' ratings of interest level, believability, and informational value? -How does advertising format affect the viewer's ability to correctly identify the Sponsor? -How does advertising format affect the evoked sets and choice criteria of viewers? CHAPTER II THEORETICAL FRAMEWORK Introductory Comment Communication is so pervasive and so much a part of our lives that we readily take it for granted, yet it is a major determinant of much of our behavior; out attitudes, opinions, moods, predispositions, and cognitions are the product of myriad communications. In an affluent, commercial, and consumption oriented world, a large number of the communications that reach the consumer's senses arrive in the form of advertising messages. A greater understanding of the nature and purposes of comparative advertising can be obtained by reviewing previous research in several areas. Information Processing There is little doubt that the issues of (consumer) information provision, acquisition, and utilization will be major problems confronting advertisers and policy makers for the remainder of this decade, yet there are at least three perspectives regarding the array of information confronting 12 13 1 The first is that of the consumer advocate the consumer. who argues that the information ought to be there because the consumer has a moral, ethical, and legal "right to know;"2 the use the consumer makes of the information is peripheal to the issue. The second perspective, that the more information consumers have, the better decisions they can make (it is implicitly assumed that consumers would use information if available and that information is processed in a uniform manner by consumers), is often adopted by public policy 3 a Federal Trade Commission priority, for example, makers; appears to be to get more facts and information into advertising.4 Proposals for information provision can be viewed as reflections of an increased public policy atten- tion to the character of the competitive environment as it 1John A. Howard and James Hulbert, Advertising and the Public Interest, (Chicago: Crain Communications, Inc., 1973L pp. 80-96; William L. Wilkie, "Analysis of Effects of Information Load," Journal of Marketing Research, (November 1974), pp. 462-466;’William L. Wiikie andTDavid M. Gardner, "The Role of Marketing Research in Public Policy Decision Making," Journal of Marketing, (January 1974), pp. 38-47. 2It is this issue of the "right to know" which serves as the doctrine of Howard and Hulbert's "full and affirma- tive disclosure." 3Wilkie and Gardner, p. 40. 4"FTC Priority is Getting Full Info, Facts in Ads, Engman Tells Lawyers," Advertising Age, (August 19, 1974), p. 4. 14 affects purchase decisions by consumers. The basic tenent of the Federal Trade Commission's policy on comparative advertising, for example, is that it provides the consumer with more information about competing products, and this, in the long run, is good for consumers in making informed choices; more precisely, such advertising possesses the potential of creating clearer consumer perceptions of brands by enabling consumers to more easily determine along which product dimensions two or more brands differ, and the extent of any such differences. The third perspective is one stemming from the behavioral sciences, especially that area dealing with human processing of information. There is substantial research evidence to suggest that there are finite limits to the ability of individuals - of consumers - to receive,process, and remember information during any given period of time.5 Since the amount of information generated by the senses is far in excess of the capacity of the brain to receive, 5Steven H. Chaffee and Jack M. McLeod, "Consumer Decisions and Information Use," in Consumer Behavior: Theoretical Sources, S. Ward and T. S. Robertson, eds., (Englewood Cliffs, N.J.: Prentice-Hall, 1973), pp. 385-415; Jacob Jacoby, Donald E. Speller, and Carol A. Kohn, "Brand Choice Behavior as a Function of Information Load," Journal of Marketing Research, (February 1974), pp. 65-67; Jacob Jacoby, Donald E. Speller, and Carol Kohn Berning, "Brand Choice Behavior as a Function of Information Load: Replication and Extension," Journal of Consumer Research, (June 1974), pp. 40-41; James Hulbert,!“Information Processing Capacity and Attitude Measurement," Journal of Marketing Research, (February 1975), p. 106. 15 process, and remember, the individual selectively processes (filters) incoming data, reducing it to a more manageable size and order. This filtering does not appear to be a random process; rather it seems to operate in a manner consistent with the individual's motives, values, attitudes, and physiological needs.6 Tannenbaum, for example, cited several examples in support of a general phenomenon in mass communication situations which he termed the "indexing process." He observed that a particular cue or cue combination, called an "index," within the total message structure could raise the threshold of attention for the message and guide its selection over other messages.7 This indexing process suggests, then, that when a comparative advertisement names or prominently displays a competitor brand which has meaning or Significance for the message recipient (such as a brand he is presently using or a well known industry leader), that the named brand may function as an "index," raising the attention level for the message. A cornerstone of communications research has long been the notion of selective exposure, the idea that people pre- fer exposure to communications that support their 6Homer E. Spence and James F. Engel, "The Impact of Brand Preference on the Perception of Brand Names: A Laboratory Analysis," in Research in Consumer Behavior, David T. Kollat, Roger D. Blackwell, and James Engel, eds., (New York: Holt, Rinehart 8 Winston, 1970), p. 61. 7Percy H. Tannenbaum, "The Indexing Process in Communication," Public Opinion Quarterly, (Summer 1955), pp. 292-302. 16 predispositions and avoid incongruent messages (although Sears and Freedman have argued that the evidence does not always Show selective exposure but simply the greater availability of consonant communications to most of the 8 It has been found rather pe0ple most of the time). consistently, for example, that individuals with strong brand preferences exhibit a heightened awareness and sensitivity to advertisements which support their prefer- ences, whereas advertisements for nonpreferred products are either ignored (selective exposure), misperceived (selective distortion), or forgotten (selective retention).9 The rationale usually advanced for the phenomenon of supportive exposure is dissonance theory: selective expo- sure to information which adds consonant elements, or weakens dissonant elements, avoids a dissonance confrontation. The theory of cognitive dissonance claims that the individual - the consumer - strives toward consistency within himself. His attitudes, values, and beliefs are ordered into clusters that are internally consistent, and consistent with his behavior. Whenever cognitions are dissonant or in conflict with each other or with behavior (or with information), tension is created, which motivates the person to a certain course of action directed toward 8David 0. Sears and Johnathan L. Freedman, "Selective Exposure to Information: A Critical Review," Public Opinion Quarterly, (Summer 1967), p. 212. 9Spence and Engel, p. 61. 17 reducing that tension (the theory does not specify the mode of dissonance reduction, but seeking and recall of consonant information, avoidance of dissonant information, perceptual distortion, opinion change, attitude change, and behavioral change are some of the common means of dissonance reduction).10 Dissonance may be aroused after making an important and difficult decision, after being coerced to say or do some- thing which is contrary to private attitudes, opinions, or beliefs, or, after being exposed to discrepant information.11 Discrepant information, interestingly, appears to be an inherent characteristic of comparative advertising. The typical comparative advertisement confronts the message recipient with information that contradicts what he believes about the attributes of his preferred brand; it states directly or by implication that an unchosen competitor brand is better than the one he is now using. Thus the user of the compared-to brand is confronted with an inconsistency which he must reconcile. While it is difficult to predict the boundary of acceptable discrepancy, media messages and appeals which 10Harold H. Kassajian and Joel B. Cohen, "Cognitive Dissonance and Behavior," California Management Review, (Fall 1965), pp. 55-58; Sadaomi Oshikawa, "Can Cognitive Dissonance Theory Explain Consumer Behavior?" Journal of Marketing, (October 1969), pp. 44-45; John A. Howard and Jagdish N. Sheth, The Theory of Buyer Behavior, (New York: John Wiley & Sons, Inc., 1969(, pp. 180-183. IIOshikawa, p. 44. 18 deviate substantially from the recipeint's own predispositions and norms of behavior are likely to be converted, in the event of exposure, to positions more nearly compatible with his own. Similarly, there is a substantially reduced tendency to accept a discrepant message when the source is of moderate of low credibility.12 Wilkie and Farris suggest that source credibility has two components: the competence of the source with respect to the particular topic, and the presence or absence of manipulative intent.l3 One might assume that a comparative advertisement for one brand would not normally be viewed by consumers as a very credible source of information about the competing brands mentioned in the advertisement since they would be likely to perceive the source as partial, biased, and as having manipulative intent (even though it may be seen as competent). Conversely, by the very fact that an advertiser is permitted to compare his brand with an identified competitor, he and his message may gain credibility; consumers may believe that in order to be printed or broadcast, the comparison must be "correct."14 12James F. Engel, David T. Kollat, and Roger 0. Black- well, Consumer Behavior, (Hinsdale, Illinois: Dryden Press, 2nd Edition, 1973), p. 320. 13Wilkie and Farris, p. 13. 141bid., pp. 13-14 19 Since the Span of immediate memory imposes distinct limitations on the amount of information individuals are able to receive,process and remember, buyers, in still another attempt to cope with the vast volume of product information, typically concentrate their purchases on a subset of the available brands and base their purchase decisions on those product attribute dimensions they deem most important rather than on all available information. The set of brands which the buyer actually considers when making a specific brand choice constitute his evoked set.15 It is a Simplification of the real world (which is composed of all brands available to the buyer). Purchase behavior can only occur among those brands for which the buyer has knowledge of their attributes, thus the magnitude of the evoked set sets a limit on the breadth of the buyer's choice at any point in time. Evaluative or choice criteria are those latent dimen- sions that are salient in the buyer's evaluation of a brand.16 They are product-Specific manifestations of the buyer's personality and his stored information amd experience (the 15Howard and Sheth, p. 98; Brian Milton Campbell, "The Existence of Evoked Set and Determinants of its Magnitude in Brand Choice Behavior," in Buyer Behavior: Theoretical and Empirical Foundations, John A. Howard and Lyman E. Ostlund, eds., (New York: Alfred A. Knopf, 1973), p. 243. 16Howard and Sheth, p. 118; Engel, Kollat, and Blackwell, pp. 248-249. 20 buyer must have some knowledge of the alternative brands before Specifying those dimensions which are important in his brand evaluation). Their purpose is to serve as standards or guidelines against which alternative brands are compared and evaluated. The buyer's evaluation of the brand's potential to satisfy his wants becomes his attitude toward that brand.17 By virtue of being the principal agent in the commercial environment, the advertiser can influence the evoked sets, choice criteria, and attitudes of potential 18 buyers. By comparing brands on several dimensions, the IcompaFative advertiser attempts to convince the consumer of the relative importance of various choice criteria and to provide the buyer with enough evaluative characteristics of the brand to enable him to adequately order it as one of several alternatives in his evoked set. To the extent that the consumer accepts this brand into his evoked set and accepts the conclusions of relative performance capabilities, as indicated by the comparison, the advertiser has been successful in altering attitudes with respect to his brand. 17Howard and Sheth, p. 126. 18Baird, Thomas R. "An Experimental Study of the Effects of Information From Advertisements and Consumer Rating Publications on Criteria Listed by Consumers as Important to Product Selection," (Ph.D. Dissertation, University of Arkansas, 1977.) 21 Message Sidedness Various propaganda strategists have put forth the claim that in appealing for acceptance of any Specific belief or policy, no opposing arguments should be discussed because mentioning rivals ideas invites comparison, hesitation and doubt.19 In their classic wartime studies, Hovland, Lumsdaine, and Sheffield set forth to explore the following question: When the weight of evidence supports the main thesis being presented, is it more effective to present only material supporting the point being made, or is it better to introduce also the arguments of those opposed to the point being made?20 In research on persuasive communication, a communication that presents arguments for a given point of view without any mention of arguments for the opposing point of view or attempted refutation of them has been called "one-Sided;" a communication that presents arguments for a given point of view, and then goes on to enumerate arguments for the opposing point of view at least some of which are refuted, has been called "two-Sided."21 19Arthur A. Lumsdaine and Irving L. Janis, "Resistance to 'Counter-propaganda' Provided by One-Sided and Two-Sided 'Propaganda' Presentations," Public Opinion Quarterly, (Fall 1953), p. 311. 20Carl 1. Hovland, Arthur A. Lumsdaine, and Fred 0. Sheffield, Experiments on Mass Communication, (Princeton, N.J.: Princeton University Press, 1949), p. 201. 21Chester A. Insko, "One-Sided Versus Two-Sided Communications and Countercommunications," Journal of Abnormal and Social Psychology, (September 1962), p. 203. 22 Early research by Hovland, Lumsdaine, and Sheffield found no overall difference in the effectiveness of persuasive communications that presented only supporting arguments (e.g., one-sided) versus those that also mentioned opposing arguments (e.g., two-sided). However, further analysis by these early researchers indicated that a two-Sided communication was more effective for those who were relatively more educated and for those who opposed the position of the persuasive communication. These findings led the investigators to speculate that bias in communication would more likely be seen as biased by those who are aware of opposing arguments (those who initially Oppose the position advocated) or who would generally be more aware of arguments on both sides of the issue (relatively well-educated people).22 Lumsdaine and Janis extended the preceeding study by comparing the effects of a one-sided and a two-sided presentation when followed by a "later“ opposing communication. Evidence from their experiment revealed that when an audience was subsequently exposed to counter- propaganda arguing in favor of the opposing position, a persuasive communication which advocates a definite position on a controversial issue is more effective in the long run if it presents and discusses the opposing arguments, than if 22Hovland, et. al., pp. 201-227. 23 it presents only the arguments that support the communicator's position.23 The rationale leading to the expectation of superior effectiveness for the two-sided over the one-sided communication, as reported by Lumsdaine and Janis, differs somewhat from the earlier experiment of Hovland, Lumsdaine, and Sheffield. In the earlier experiment which was concerned with immediate effects, the explanation is as follows: If a member of an audience has a strongly held "negative" opinion and is familiar with arguments supporting this opinion, presentation of a one-sided communication supporting the positive point of view will encounter a relatively high degree of resistance. The listener's tendency to think of his own arguments will interfere with acceptance of (or even attention to) the arguments that the communicator is presenting. If, however, the communicator explicitly discusses the listener's own negative arguments, this source of interference is reduced. Moreover, the listener may be more impressed by the communication and less inclined to distrust the arguments as coming from a biased source if the communicator makes it clear that he has taken the negative arguments into account.24 (More 23Lumsdaine and Janis, p. 317. 24Ibid., p. 317. 24 recent evidence documents that the two-sided appeal may have a positive effect on perceived claim and source credibility.)25 Lumsdaine and Janis, however, explained their results as follows. Regardless of initial position, a convincing one-Sided communication presenting only positive arguments will tend to sway many members of the audience further in the direction advocated by the communicator. However, when these persons subsequently hear the Opposite point of view, also supported by cogent-sounding arguments, their opinions tend to be swayed back in the negative direction (especially if the new arguments appear to offset the previous positive arguments). But if the initial communication is, instead, a two-Sided one it will already have taken into account both the positive and negative arguments and still have reached the positive conclusion. When the listener is subsequently exposed to opposing arguments in the counter- propaganda, he is less likely to be influenced by them. He is not only familiar with the opposing point of view, but has been led to the positive conclusion in a context in which the negative arguments were in evidence. In effect, he has 25Robert B. Settle and Linda L. Golden, "Attribution Theory and Advertiser Credibility," Journal of Marketing Research, (May 1974), pp. 184-185; Kenneth J. Roering and Robert J. Paul, "The Effect of the Consistency of Product Claims on the Credibility of Persuasive Messages," Journal of Advertising, (Spring 1976), pp. 34-36. 25 been given an advance basis for ignoring or discounting the opposing communication, and, thus "inoculated," he 26. (More recent will tend to retain the positive conclusion. evidence documents that two-sided communications are more effective in protecting the message recipient from counter- propaganda and countercommunications.)27 Evidence reported by Chu supports the theoretical position that awareness of relevant arguments on the issue (including at least some opposed to the position advocated by the communication) and opposition to the position of the communication are necessary conditions for a one-sided communication to be less effective than a two-Sided one.28 Chu concluded that differential responses to one-and two- Sided communications were probably due to perceived bias in the communications (regardless of the style of the argument or initial attitudes, the effects of communication were an inverse function of the detection of bias) and not to the rehearsal of counterarguments (early researchers frequently 26Lumsdaine and Janis, p. 318. 27Stewart w. Bither, Ira J. Dolich, and Elaine B. Nell, "The Application of Attitude Immunization Techniques in Marketing," Journal of Marketing Research, (February 1971), pp. 59-60; George J. Szybillo and'Richard Heslin, "Resistance to Persuasion: Inoculation Theory in a Marketing Context," Journal of Marketing Research, (November 1973), pp. 400-403. 28Godwin C. Chu, "Prior Familiarity, Perceived Bias, and One-Sided Versus Two-Sided Communications," Journal of Experimental Social Psychology, (July 1967), pp. - . 26 expressed the belief that covert rehearsal by the message recipient of arguments opposing the position recommended by the communicator would lessen persuasion).29 While Chu's experiment supports the major line of reasoning of Hovland et. al., Jones and Brehm have suggested that there may be a somewhat more subtle underlying process to account for the differential effectiveness of one-and two-sided communications. Their theory asserts that when a person feels free to adopt or reject any of several positions on an issue, pressure to adopt a particular position threatens his freedom and arouses a motivational state called "psychological reactance.“30 Since a one-sided communication is normally seen as exerting greater pressure to adopt a particular position than a two-Sided communica- tion, it will tend to produce more reactance, and consequent resistance (reactance leads the individual to try to restore the threatened freedom by resisting the pressure to adopt the particular position), than will a two-Sided one. However, where the individual is quite ignorant about the 291bid. 30The necessary conditions for the arousal of reactance are: (1) that the individual feel free to adopt a position on either Side of an issue, (2) that this freedom have some importance to him, and (3) that the communication be perceived as a threat to his freedom to adopt or reject one of the possible positions. (Russell A. Jones and Jack W. Brehm, "Persuasiveness of One-and Two-Sided Communications as a Function of Awareness There Are Two Sides," Journal of Experimental Social Psychology, (January 1971), p. 49;) 27 issue and does not know whether or not there really are two sides, a one-Sided communication will be relatively effective since it provides a less equivocal position than a two-sided communication that informs the message recipient that there are two sides and pressures him to adopt only one.31 Jones and Brehm concluded from their experiment that a one-sided communication would be reduced in effectiveness, relative to a two-sided communication, to the extent that the audience is aware that there are two plausible positions on the issue (even though it is not necessary that he know any specific arguments or facts). Furthermore, while a predisposition to favor one side could increase the importance of being able to reject a position on the other side, it is not necessary for the arousal of reactance that the individual be initially opposed to the position advocated by the communication.32 Thus, the objective evidence indicates that the two- sided argument, under certain conditions, can be a very effective message appeal. Generally, the evidence indicates (and it is widely accepted) that a two-Sided appeal is more effective at getting across its message for those who are familiar with the issue, initially opposed to the position of the argument, better educated, and where 31Ioid. 321oid., p. 55. 28 counterpropaganda is subsequently encountered. Extension of these findings to the field of advertising, however, requires caution; there are elements in the (typical) advertising environment that are quite different from these experiments: the time available for the presentation of the message is seriously reduced, there is reduced credi- bility of the communication environment, and exposure, attention, perception and learning are commonly unmotivated and often selective.33 Comparative Dimensions Those familiar with communications research will see distinct Similarities between one-sided versus two-Sided and comparative communications. AS noted earlier, one-Sided communication involves presenting material which supports only the communicator, while two-Sided communication involves presenting material which both supports and is opposed to the communicator. Comparative advertising was previously defined as advertising that compares two or more Specifically named or recognizably presented brands of the same generic product or service class, and makes such a comparison in terms of one or more specific product or service attributes. Both one-and two-Sided communications, therefore, may be comparative or non-comparative. A 33Alan G. Sawyer, "The Effects of Repetition of Refutational and Supportive Advertising Appeals," Journal of Marketing Research, (February 1973), p. 24. 29 framework for conceptualizing and evaluating argumentative structures which recognizes this interaction, and one that will be used throughout this inquiry, has been suggested by Mazis (Figure 2). Although one-versus two-Sided communications have been used in numerous studies, there has been relatively little attention devoted to the comparative, or non-comparative, aspects of communications. This indifference to the comparative aspects of one-and two-sided communication by social psychologists is largely explained by the issues studied; rather than commercial messages, nearly all of the reported research has concentrated on speeches and essays advocating a particular position on some issue of current importance.34 Under these message conditions, negative features of the communicator's position are closely related to positive aspects of the opposing view (i.e., negative feelings toward one side imply a positive view toward the other side). The assumption has frequently been made that positive statements about the communicator's position (PAl) will produce equivalent effects as negative statements about the opposing Side (NA2); arguments, detrimental to the communicator's belief (NAl) are presumed analagous to 34e.g., an early end to World War II after the surrender of Germany (Hovland, et. al., 1949), the production of atomic bombs by Russia (Lumsdaine and Janis, 1953), legal cases (Insko, 1962; Jones and Brehm, 1970), the creation of an international free trade zone in Taiwan (Chu, 1967), and the guarantee of an annual minimum cash income (Koehler, 1972). 30 One-Sided Two-Sided Non-Comparative PAl NAl, PAl (NAZ) (PAZ’ NAZ) Comparative PAI’ NA2 PAl, PA2 (NAl, NAZ) P42, NAZ, 9A1 Where: PA represents a positive argument NA represents a negative argument Subscripts indicate which antagonist is represented (1 represents the communicator/advertiser) (2 represents an opposing side/competitor) Parentheses indicate an argument unlikely to occur FIGURE 2 Framework for Conceptualizing Argumentative Structures35 35Michael B. Mazis, "A Framework for Evaluating the Effects of Comparative Advertising," (Paper presented to the American Marketing Association Educators' Conference, Rochester, N.Y., August 1975), p. 3. 31 arguments advocating the antagonist's view (PA2).36 Therefore, little attention to the comparative dimension was necessary in previous research. Brand advertising, on the other hand, often involves a number of competing brands or positions, and a favorable attitude toward one brand does not preclude having a favorable attitude toward competing brands; extolling the virtues of the advertised brand may be disassociated from disparaging opposing brands. Not surprisingly, then, there is mixed support for the aforementioned message Sidedness findings in published studies using advertising messages. Faison, consistent with Hovland, et. al.'s research, found two-Sided advertising communications more effective in influencing attitudes of those initially opposed to the point of view presented in the commercials (i.e., users of competitive products) and of those with higher intelligence. Additionally, Faison concluded that the relative amount of influence of both one-and two-Sided advertising communication seemed to be related to the type of product being advertised (the commercial concerning the least expensive product, floor wax, produced a much greater change in attitudes than did the commercials for the most expensive product, (automobiles) and to the knowledge of the consumer (for both one-and two-Sided commercials, the more knowledge a person had about 36Mazis, p. 5. 32 a given subject, the less influential a particular communication was in altering his opinion).37 Unfortunately, it is not clear whether the differences between the treatments were due to the presentation of two sides or comparison against competitive brands (e.g., the one-sided versus the two-sided variable may have been confounded with a comparative variable). Golden concluded that there were no meaningful differences in the effectiveness of one-Sided comparative and non-comparative messages.38 However, Similar to the Faison study, Golden's commercials were presented to subjects without the usual program environment, thereby increasing the likelihood of subjects supplying normative beliefs about which commercials they should prefer and limiting the generalizability of the findings. Still another difficulty in interpreting Golden's findings is the weak nature of the experimental manipulation; the failure of respondents to believe any of the claims made, whether in comparative or non-comparative form may have contributed to the retention of the null hypotheses. 37Edmund Winston Jordan Faison, "Effectiveness of One- Sided and Two-Sided Mass Communications in Advertising," Public Opinion Quarterly, (Fall 1961), p. 469. 38Linda L. Golden, "Consumer Reactions to Direct Brand Comparisons in Advertisements," (Ph.D. Dissertation, University of Florida, 1975). 33 Although there was some tendency for greater sponsor misidentification under the comparative treatment, Mazis concluded that the major impact of one-Sided comparative advertising was on the evaluation of the brand being attacked (attitude change was manifested principally lower ratings for the brand which was disparaged by the comparative advertisement) and not on the rating of the sponsoring brand.39 Prasad, in a test of a (one-sided)comparative (print) advertisement in relation to its "Brand X" counterpart, found that although a comparative advertisement can enhance message recall, it can also result in some loss of effectiveness from consumer's perception of low credibility of its claims. The results of the experiment indicated that, on the whole, the message recall effectiveness of the comparative advertisement was higher than that of its "Brand X" counterpart, and that the perceived credibility ratings of the product superiority claim of the comparative advertisement were lower among subjects who had a prior preference for the competitive brand named in the advertise- 4o ment than among others. Contrary to the opinion held by 39Mazis, pp. 17-21. 40Kanti V. Prasad, "Communications-Effectiveness of Comparative Advertising: A Laboratory Analysis," Journal of Marketing Research, (May 1976), pp. 132-135. 34 many advertising people, no evidence was found that a comparative advertisement was more effective than its "Brand X" counterpart in projecting for the sponsoring brand an image of competitive proximity to the industry leader, nor did the experiment produce any evidence of selective recall phenomenon among subjects whose preferred brand was attacked in the comparative advertisement to which they were exposed.41 Studies of comparative and non-comparative (television) commercials by some syndicated testing services (Burke, McCollum/Spielman, Gallup & Robinson) and by the advertising agency Ogilvy & Mather found no differences in the "persuasiveness" of the two types of commercials.42 Additionally, the Ogilvy & Mather study indicated that comparative advertising makes consumers more aware of competitors, results in lower belief of claims and increased misidentification and confusion.43 Because Ogilvy & Mather sought current, state-of-the—art, finished work (in three product categories) with identical creative strategies, it 411bid. 42Andrew G. Kershaw, "For and Against Comparative Advertising - Against," Advertising Age, (July 5, 1976), pp. 25-26; Philip Levine, "Commercials that Name Competing Brands," Journal of Advertising Research, (December 1976), 10. 43"Comparative Ads Ineffective: O & M Study," Advertising Age, (October 13, 1975), p. 16. 35 didn't have much latitude in picking the commercials it did test, and it appears likely that the comparative variable may have been confounded with a message sidedness variable. Furthermore, Ogilvy & Mather's particularly negative viewpoint it had in mind when it designed and completed the study didn't exclude veiled (Brand X) comparatives from the test (i.e., one veiled comparative was included as a non-comparative test commercial, even though it is reasonable to suspect that a "Brand X" type commercial is perceived by consumers as comparative). Although traditional levels of significance were not met, Sellers concluded that comparative advertising can be more effective when some product interest exists, when more informative advertisements are used, and when consumers depend less on personal sources of information and have generally positive perceptions of advertising.44 Abramson discovered that comparative advertising can have a negative impact on the attitudes of brand loyal customers, however, defensive measures, such as refutational advertising in combination with supportive 44Sellars, Ronald Kay. "A Study of the Effectiveness of Comparative Advertising for Selected Household Appliances," (Ph.D. Dissertation, The Louisiana State University and Agricultural and Mechanical College, 1977). 36 advertising, can reduce the negative attitude change brought about by a comparative attack.45 In a study of comparative print (i.e., magazine) advertisements, Williams concluded that comparative advertisements provide no Significantly different influences on the attitudes of readers, toward the advertised brand, than do non-comparative advertisements. While the comparative advertisement Showed some strength in gaining brand awareness and recall, the non-comparative advertisement was better than the comparative advertisement in achieving brand preference for the advertised brand. Furthermore, the experiment suggested that using comparative advertising could lead to greater brand preference for named competing brands.46 Although these studies represent a worthwhile step toward providing empirical data on comparative advertising, these findings cannot be accepted as unequivocal. Consideration of the comparative variable is extremely important if message research is to be applicable to the 45Abramson, Joseph. "Comparative Advertising, Inoculation Theory, and the Prevention of Attitude Change Among Brand Loyal Customers," (Ph.D. Dissertation, The Louisiana State University and Agricultural and Mechanical College, 1977). 46Williams, Terrell G. "An Experiment in the Effects of Comparative Advertising on Brand Awareness, Comprehension, and Preference," Southern Business Review, (Spring 1978), 40. 37 marketing environment, therefore, additional research, under more closely controlled manipulation and of a more realistic nature, seems warranted. CHAPTER III THE RESEARCH DESIGN Introductory Comment AS discussed earlier, the objective of this research study is to further our understanding of comparative advertising effectiveness by examining the effects achieved by comparative versus more traditional advertising formats. In this chapter, an overall program of research is developed for organizing and conducting the research in order to bring empirical evidence to bear on the related research questions. Before detailing this research program, however, the research questions are stated in the form of hypotheses suitable to be empirically tested. Hypptheses Consistent with the objective of analyzing the effects of comparative advertising, hypotheses were developed along S x majordimensions: (1) interest level, (2) believability, (3) informational value, (4) sponsor identification, (5) evoked set, and (6) choice criteria. 38 39 INTEREST LEVEL H Comparative advertisements will be rated as being 1. more interesting than non-comparative advertisements. H2: Two-sided comparative advertisements will be rated as being more interesting than one-Sided comparative advertisements. BELIEVABILITY H3: Comparative advertisements will be rated as being more believable than non-comparative advertisements. H4: Two-Sided comparative advertisements will be rated as being more believable than one-Sided comparative advertisements. INFORMATIONAL VALUE H5: H6: Comparative advertisements will be rated as being more informative than non-comparative advertisements. Two-Sided comparative advertisements will be rated as being more informative than one-sided comparative advertisements. SPONSOR IDENTIFICATION H7: H8: Comparative advertisements will encounter more Sponsor misidentification than will non-comparative advertisements. Two-sided comparative advertisements will encounter more Sponsor misidentification than will one-Sided comparative advertisements. EVOKED SET H9: H10: The evoked sets of those exposed to comparative advertisements will be more consistent with the brands named in the comparisons than will the evoked sets of those exposed to non-comparative advertisements. The evoked sets of those exposed to two-Sided compara- tive advertisements will be more consistent with the brands named in the comparisons than will the evoked sets of those exposed to one-sided comparative advertisements. 4O H11: Owners of named competing brands are more likely to admit the sponsored brand to their evoked set than are owners of brands not mentioned. CHOICE CRITERIA H12: The choice criteria of those exposed to comparative advertisements will be more consistent with the brand attributes used in the comparisons than will the choice criteria of those exposed to non-comparative advertisements. H13: The choice criteria of those exposed to two-Sided comparative advertisements will be more consistent with the brand attributes used in the comparisons than will the choice criteria of those exposed to one-sided comparative advertisements. H14: Owners of named competing brands are more likely to accept the brand attributes used in the comparisons as their choice criteria than are owners of brands not named. Methodology Experimental television advertisements of four types were prepared to represent four treatment conditions - exposure to a one-sided non-comparative advertisement, a one-Sided comparative advertisement, a two-sided non- comparative advertisement, or, a two-sided comparative advertisement. Structurally, the four advertisements were identical. Each 60-second color advertisement utilized the same brand claims, order of presentation of claims, studio setting, equipment, and personnel. The only element of difference between the four mock advertisements was the format in which the brand claims were presented - two—sided advertisements disclaimed superiority on some product attri- butes and comparative advertisements identified competitor 41 brands. (A complete audio/video script for each of the four advertisements appears in Appendix A.) The four mock advertisements were for an existing brand of digital watch. Pretests indicated that digital watches could serve as an appropriate product for application of the experimental manipulations - brand preferences were not clearly formed (although Bulova and Timex were perceived as industry leaders while Litronix was relatively unknown), it was viewed as a significant purchase, and, relatedly, buyers normally made comparisons along several dimensions before purchasing. Since many people in advertising believe that comparative advertising works best when an underdog brand attacks an (established) industry leader,1 Litronix was selected to be the sponsoring brand, while Bulova and Timex were the competitor brands identified in the comparative advertisements. The brand claims made in all four advertisements (low price, electronics excellence, one-year warranty, IO-second accuracy, and ease of mainten- ance) were extracted from pretest data, Litronix sales promotion material, and prior Litronix advertisements. To minimize the possibility of subjects having a predominantly evaluative response set, such as has been the case when nothing but the advertisement was presented for 1"Underdog Advertiser Wins in 'Naming Names': 8800," Advertising Age, (March 10, 1975), p. 56; Aimee L. Morner, "It Pays to Knock Your Competitor," Fortune, (February 13, 1978), 104. 42 their evaluation, the experimental advertisements were edited into a television program. The television program selected for use in this study was a half-hour (color) segment of the Dinah Show. The particular show actually used was entitled "Dinah and the Gold Record Winners," and was taped off the air, complete with its commercials (a brief outline of the program character and sequence is presented in Appendix B). It contained no controversial issues which could arouse strong feelings or affect responses, nor did it supply any normative beliefs about the product or brands employed in the mock advertisements. The only modification made to the program was the substitution, via random assignment, of an experimental advertisement in the third commercial time spot (cost and sample size limitations precluded rotation of the experimental advertise- ment to all five commercial time spots). The experimental design used in the administration of the experimental manipulations is depicted in Figure 3. Treatment (program) types were randomly assigned to each of the four subject groups. Subjects (125 junior-level students enrolled in an introductory marketing course at Michigan State University in the spring of 1976) were first told that part of the classroom procedure for that day would be a group exercise in television communications (the assigned topic for this class session, "mass communication models and processes," provided justification for the television program to which the subjects were to be exposedL 43 Before Experimental After Measurement Variable Measurement OSNC OSC TSNC TSC 1 No Yes No No No Yes 2 No No Yes No No Yes Group No No No Yes No Yes 4 No No No No Yes Yes Where: OSNC represents a one-sided non-comparative advertisement OSC represents a one-Sided comparative advertise- ment TSNC represents a two-Sided non-comparative advertisement TSC represents a two-Sided comparative advertise- ment FIGURE 3 Experimental Design 44 Subjects were then randomly assigned to one of the four treatment groups and assigned to adjoining classrooms (with seats arranged to minimize verbal interaction). The following directions were then read to them. Today you will be watching a current television program that was taped off the air, in its entirety, at an earlier data. It is not necessary that you attempt to memorize any of the program or commercial content. A brief discussion, involving all groups, will follow immediately after watching the program. Following these instructions, subjects were shown the randomly assigned television program (to avoid the sensitizing effects of a pretest no pre-exposure measure- ments were taken). Although the instructions stated that a brief group discussion would follow (a ploy that pretests indicated was effective in discouraging memorization of program and commercial content), a questionnaire relating to the focus of the research was administered instead. The Qgestionnaire The postexposure questionnaire consisted of 18 multiple choice and (seven-point) Likert-type questions. These questions were designed to elicit accurate and uniform responses. A preteSt of the survey instrument served to identify ambiguities obstructing measurement of the variables under test or prohibiting data interpretations or comparisons. The questionnaire consisted of three parts. Instruc- tions on the cover page requested subjects to respond 45 anonymously to a series of questions about the program and its commercials. The second part consisted of 14 questions designed to provide measures of group Similarity, program and sponsor identification, interest, information and believability perceptions, claim recall, brand attribute saliency, product and brand ownership, and purchase intentions. The four questions comprising the third part of the questionnaire sought demographic data about the respondents. A sample of the questionnaire appears in Appendix C. Non-responses to the questions dealing with advertising themes and sponsor identification were recorded as incorrect Unintelligible answers resulted in the elimination of three completed questionnaires, reducing the data base to 125 respondents. Data Analysis Aside from the usual statistics and contingency tables, the principal statistical technique employed was the Kolmogorov-Smirnov two-sample test. Earlier in this chapter, a series of hypotheses were advanced stating that there were Significant differences expected from the administration of specified experimental manipulations. The (two-tailed) Kolmogorov-Smirnov two- sample test is sensitive to any kind of differences (central tendency, dispersion, skewness, etc.) in the distributions from which the samples were drawn, and the (one-tailed)test 46 can be used to decide whether or not the values of the population from which one of the samples was drawn are stochastically larger than the values of the population from which other samples were drawn.2 Data analysis was accomplished through use of the SPSS procedures "CROSSTABS" and "NPAR TESTS". ConcludinggComment In Summary, this chapter presented the research design for a large-scale laboratory-type experiment. The structure and strategies employed were deliberately and specifically conceived and designed to simulate a realistic environment, to offer a high degree of control over experimental and measurement procedures, and to bring evidence to bear on the research questions. 2Sidney Siegel, Nonparametric Statistics, (New York: McGraw-Hill Book Company, 1956), p. 127. CHAPTER IV RESULTS OF THE RESEARCH Introductorngomment The objective of the research design, which was addressed in detail in the preceeding chapter, was to allow for the examination of variations in advertising format at six levels of analysis: interest level, believability, informational value, Sponsor identification, composition of the evoked set, and statement of choice criteria. These areas of analysis provide the organizing framework for this chapter, which concludes with a review of selected behavioral and demographic characteristics of the sample. Interest Level For the first research dimension of "interest level" it was hypothesized that: (H1) comparative advertisements will be rated as being more interesting than non-comparative advertisements, and, (H2) two-Sided comparative advertise- ments will be rated as being more interesting than one- sided comparative advertisements. The research findings relating to H1 are presented in Table 1. They clearly Show (Z = 1.7062, p = .0030) that 47 48 TABLE 1 Interest Level and the Comparative Dimension number meana standard deviation one-sided two-sided row total 30 32 62 non-comparative 5.1000 4.6250 4.8548 .8449 .9070 .9026 34 29 63 comparative 4.0588 4.2414 4.1429 1.3244 1.0907 1.2162 Maximum (+) difference .4176 .1864 .2962 Kolmogorov-Smirnov Z 1.7359 .7978 1.7062 1-tailed p .0024 .2800 .0030 aratings can range from 1 to 7, where "1" represents the most desirable interest rating and "7" represents the least desirable interest rating. 49 comparative advertisements were rated as being more interesting than their non-comparative counterparts. This effect, of improved interest level ratings, was observed for both one-sided and two-sided comparative presentations, however, only the one-Sided treatments displayed a difference that was large enough to be statistically significant (0 = .4176, z = 1.7359, p = .0024).1 The research findings presented in Table 2 indicate that there were no statistically significant differences between the interest level ratings of one-sided and two- Sided comparative advertisements (D = .1410, Z = .6263, p = .4563). Contrary to H2, two-sided comparative advertisements received less favorable interest level ratings than one-sided comparative advertisements; two- sided presentations registered superior interest level ratings only when presented in the non-comparative format. The results of this research, then, provides evidence to support H1, but not to support H2. Comparative advertisements, although not necessarily perceived as being more interesting than most television advertisements (i.e., Y > 4), were rated as being more interesting than their non-comparative complements. Furthermore, this interest effect was most pronounced when the comparative dimension was presented singularly, not in combination with the two- sided format. 1A .05 level of significance will be used to evaluate all research findings for all hypotheses. 50 TABLE 2 Interest Level and the Sidedness Dimension number meana standard deviation non- row comparative comparative total 3O 34 64 one-sided 5.1000 4.0588 4.5470 .8449 1.3244 1.2335 32 29 61 two-sided 4.6250 4.2414 4.4430 .9070 1.0907 1.0087 Maximum (+) difference .2688 .1410 .1355 Kolmogorov-Smirnov 2 1.1305 .6263 .8073 1-tailed p .0776 .4563 .2716 aratings can range from 1 to 7, where "1" represents the most desirable interest rating and "7" represents the least desirable interest rating. 51 Believability For the second research dimension it was hypothesized that: (H3) comparative advertisements will be rated as being more believable than non-comparative advertisements, and, (H4) two-Sided comparative advertisements will be rated as being more believable than one-sided comparative advertise- ments. Similar to the findings on interest level ratings, the research findings on believability ratings, presented in Table 3, Show that comparative advertisements were rated as being more believable than their non-comparative equivalents (A = 1.8035, p = .0015), and that this effect of enhanced believability ratings for comparative advertisements was most significant for one-sided treatements (D = .3961, Z = 1.6498, p = .0043). Those individuals subjected to two-sided presentations assigned more favorable believability ratings to the comparative presentation, however, these ratings did not produce intergroup differences that were statistically significant (0 = .2425, Z = 1.0164, p = .1267) The research findings relating to H4 appear in Table 4. They reveal that while the two-sided comparative advertise- ment received the best believability rating of any advertisement tested (7 = 3.1379), its believability rating was not Significantly different from the believability rating of the one-Sided comparative (D = .0994, Z = .4618, p = .6528). Thus, H4 was not supported; two-sided compara- tive advertisements did not receive significantly better Believability and the Comparative Dimension 52 TABLE 3 number meana standard deviation one-sided two-Sided row total 30 32 62 non-comparative 4.5333 3.9063 4.2097 .9371 1.0583 1.0423 34 29 63 comparative 3.3824 3.1379 3.2698 .9852 .9533 .9706 Maximum (+) difference .3961 .2425 .3136 Kolmogorov-Smirnov Z 1.6498 1.0164 1.8035 l-tailed p .0043 .1267 .0015 aratings can range from 1 to 7, where "1" represents the most desirable believability rating and "7" represents the least desirable believability rating. 53 TABLE 4 Believability and the Sidedness Dimension number meana standard deviation non- row comparative comparative total 30 34 64 one-sided 4.5333 3.3824 3.9223 .9371 .9852 1.1170 32 29 61 two-sided 3.9063 3.1379 3.5411 1.0583 . .9533 1.0735 Maximum (+) difference .3125 .0994 .1821 Kolmogorov-Smirnov Z 1.3027 .4618 1.0678 1-tailed p .0336 .6528 .1022 aratings can range from 1 to 7, where "1" represents the most desirable believability rating and "7" represents the least desirable believability rating. 54 believability ratings than one-Sided comparative advertisements. The results of the research on the believability dimension, then, provide evidence to support H3, but not to support H4. While only the one-Sided non-comparative advertisements failed to be rated as more believable than 2 comparative advertisements most television advertisements, were consistently rated as being more believable than their non-comparative counterparts. The believability rating effects resulting from a change in message sidedness, once the advertisement was in its comparative form, were not statistically significant. Informational Value Analogous to the hypotheses developed for the research dimensions of interest level and believability, it was further hypothesized that: (H5) comparative advertisements will be rated as being more informative than non-comparative advertisements, and, (H6) two-Sided comparative advertise- ments will be rated as being more informative than one- Sided comparative advertisements. The research findings presented in Table 5 clearly Show that comparative advertisements were consistently and unequivocally rated as being more informative (Z = 1.9796, 2A condition that conveys treatment-response validity (i.e., conventional advertisements were not rated as being more interesting than conventional advertisements). 55 TABLE 5 Informational Value and the Comparative Dimension number meana standard deviation one-sided two-sided row total 3O 32 62 non-comparative 4.4333 4.0000 4.2097 .9353 1.0473 1.0104 34 29 63 comparative 3.1765 2.9655 3.0794 1.0580 1.1797 1.1115 Maximum (+) difference .4216 .3136 .3451 Komogorov-Smirnov Z 1.7516 1.2938 1.9796 l-tailed p .0022 .0352 .0004 aratings can range from 1 to 7, where "1" represents the most desirable informational value rating and "7" represents the least desirable informational value rating. 56 p = .0004) than their non-comparative complements. One- sided treatments, once again, supplied the most conspicuous and significant differences (0 = .4216, Z = 1.7516, p = .0022), however, for the first time, the two-Sided treatments also exhibited statistically Significant differences in informational value ratings for comparative and non-comparative treatments (D = .3136, Z = 1.2938, p = .0352). Despite the fact that the two-sided comparative treatment received the most favorable informational value rating (X = 2.9655), the research findings presented in Table 6 disclose no statistically significant differences in the informational value ratings of one-sided and two- sided comparative advertisements (D = .0801, Z = .3856, p = .7428). The results of this research, then, provide sufficient evidence to accept H5, but not to accept H6“ The research results reveal that not only were comparative advertisements rated as being more informative than most television advertisements (i.e., 7 < 4), but they were also rated as being more informative than their equally information ladened non-comparative counterparts. Since all test advertisements stressed the same selling points, it must be concluded that the comparative format presents the informa- tion to the viewer in either a more understandable or usable form. Interestingly, the condition of information overload was not reflected in the ratings; the two-sided comparative, 57 TABLE 6 Informational Value and the Sidedness Dimension number mean6 standard deviation non- comparative comparative row total 30 34 64 one—sided 4.4333 3.1765 3.7664 .9353 1.0580 1.1784 32 29 61 two-Sided 4.0000 2.9655 3.5082 1.0473 1.1797 1.2196 Maximum (+) difference .2167 .0801 .1017 Kolmogorov-Smirnov Z .9256 .3856 .6183 l-tailed p .1802 .7428 .4655 aratings can range from 1 to 7, where "1" represents the most desirable informational value rating and "7" represents the least desirable informational value rating. 58 the most complex and lengthy message, received the highest informational value rating. Sponsor Identification For the research dimension of Sponsor identification, it was hypothesized that: (H7) comparative advertisements will encounter more sponsor misidentification than will non-comparative advertisements, and, (H8) two-sided comparative advertisements will encounter more sponsor misidentification than will one-Sided comparative advertisements. The research findings presented in Table 7 indicate that those individuals exposed to two-sided comparative advertisements experienced the most difficulty in correctly identifying the sponsor of the advertisement (X = .7931), however, the unusually high sponsor identification scores reported for all treatments precluded the development of any intergroup differences of statistical significance. A Similar result is observed in Table 8, where the sponsor identification scores for one-sided and two-sided comparative advertisements are nearly identical, resulting in no statistically significant difference between these treatments (0 = .0010, Z = .0726, p = .9895). The results of the research on Sponsor identification provide no evidence to support either H7 or H3. The ability of the viewer to correctly identify the sponsor of an Sponsor Identification and the Comparative Dimension 59 TABLE 7 number meana standard deviation one-sided two-sided row total 30 32 62 non-comparative .8000 .9063 .8548 .4068 .2961 .3551 34 29 63 comparative .7941 .7931 .7937 .4104 .4123 .4079 Maximum (+) difference .0059 .1131 .0612 Kolmogorov-Smirnov Z .0921 .5120 .3924 l-tailed p .9832 .5919 .7349 u r m , w re r r aval e can ange fro O to 1 he "0" e esents an incorrect Sponsor identification and "1" represents a correct Sponsor identification. 60 TABLE 8 Sponsor Identification and the Sidedness Dimension number meana standard deviation non- row comparative comparative total 30 34 64 one-Sided .8000 .7941 .7969 .4068 .4101 .4055 32 29 61 two-sided .9063 .7931 .8525 .2961 .4123 .3573 Maximum (+) difference .1063 .0010 .0556 Kolmogorov-Smirnov Z .4911 .0726 .3606 l-tailed p .6173 .9895 .7710 6value can range from O to 1, where "0" represents an incorrect sponsor identification and "1" represents a correct Sponsor identification. 61 advertisement does not appear to be significantly affected by variations in message sidedness or comparativeness. Evoked Set For the research dimension called "evoked set" it was hypothesized that: (H9) the evoked sets of those exposed to comparative advertisements will be more consistent with the brands named in the comparisons than will the evoked sets of those exposed to non-comparative advertisements, (H10) the evoked sets of those exposed to two-sided comparative advertisements will be more consistent with the brands named in the comparisons than will the evoked sets of those exposed to one-Sided comparative advertisements, and, (H11) owners of named competing brands are more likely to admit the sponsored brand to their evoked set than are users of brands not mentioned. The research findings presented in Table 9 provide unequivocal support for H9. The evoked sets of those individuals exposed to comparative advertisements were decidely more consistent with the brands named in the comparison than were the evoked sets of those exposed to non-comparative advertisements (D = .6805, Z = 3.8543, p = .0000). The results were equally dramatic in both the one-sided and two-sided treatments, with comparative treatment viewers consistently identifying named competitors as major competitors in that market. 62 TABLE 9 Evoked Set and the Comparative Dimension number meana standard deviation one-sided two-sided row total 3O 32 62 non-comparative .8667 .8438 .8548 .7303 .7233 .7208 24 29 63 comparative 2.0294 2.1724 2.0952 .8699 .6584 .7770 Maximum (+) difference .6314 .7371 .6805 Kolmogorov-Smirnov Z 2.5892 2.9456 3.8543 1-tailed p .0000 .0000 .0000 ascore can range from O to 3, and represents the number of named competing brands identified. 63 Results appearing in the one-sided and two-Sided comparative cells of Table 10 reveal a slight tendency for those subjected to two-sided presentations to demonstrate more adherence to the named competitors, than those subjected to one-Sided presentations. The resultant inter- group differences, based on these evoked set composition scores, were not large enough, however, to be statistically significant (0 = .0974, Z = .4537, p = .6624); thus H10 should be rejected. Inspection of the data presented in Table 11 suggests that owners of named competing brands were not more likely to admit the Sponsored brand to their evoked set than were the owners of brands not mentioned (0 = .0319, Z = .1502, = .9431). The relatively infrequent appearance (i.e., xl-o < .5) of the sponsored brand in the evoked sets of all respondents exposed to comparative advertisements reveals that even the comparative format experiences difficulty when trying to position an unknown brand as a relevant competitor to well-known national brands. Thus, the results of the research on composition of the evoked set provide evidence to support H9, but not to support H10 or H11. Comparative advertisements significant- ly structured the viewers' identification of major competing brands, while the message sidedness and brand ownership dimensions displayed little or no effects. 64 TABLE 10 Evoked Set and the Sidedness Dimension number mean6 standard deviation non- row comparative comparative total 30 34 64 one-Sided .8667 2.0294 1.4844 .7303 .8699 .9919 32 29 61 two-sided .8438 2.1724 1.4754 .7233 .6584 .9593 Maximum (+) difference .0125 .0974 .0236 Kolmogorov-Smirnov Z .1122 .4537 .1816 1-tailed p .9706 .6624 .9361 ascores can range from O to 3, and represents the number of named competing brands identified. 65 TABLE 11 Evoked Set and Brand Ownership number meana standard deviation b row one-sidedb two-Sided total 12 15 27 non-owners .1650 .2000 .1851 .2983 .2851 .2934 22 14 36 owners .1818 .2143 .1946 .2914 .3038 .2992 Maximum (+) difference .0378 .0430 .0319 Kolmogorov-Smirnov Z .1684 .1812 .1502 1-tailed p .9152 .8961 .9431 avalue can range from O to 1, where "0" represents an evoked set where the sponsored brand is not identified and "1" represents an evoked set where the Sponsored brand is identified. le relates only to those treatments that named competitors, thug, only the response of those subjected to comparative treatments is reported. 66 Choice Criteria Analogous to the hypotheses that were developed for the evoked set research dimension, it was next hypothesized that: (H12) the choice criteria of those exposed to comparative advertisements will be more consistent with the brand attributes used in the comparisons than will the choice criteria of those exposed to non-comparative advertisements, (H13) the choice criteria of those exposed to two-sided comparative advertisements will be more consis- tent with the brand attributes used in the comparisons than will the choice criteria of those exposed to one-Sided comparative advertisements, and (H14) owners of named compe- ting brands are more likely to accept the brand attributes used in the comparisons as their choice criteria than are owners of brands not named. Research findings appearing in Table 12 do not indicate any statistically significant differences in the occurrence rates of named brand attributes in the choice criteria of comparative and non-comparative groups (0 = .0561, Z = .3638, p = .7674). Subjects exposed to the one-sided comparative message displayed the largest incidence of named brand attributes in their choice criteria (X = 2.6471) of any group tested. As noted above, one-sided comparative treatments recorded the highest incidence of named brand attributes in their choice criteria of any group tested, thus, it was not surprising to find that the choice criteria of those 67 TABLE 12 Choice Criteria and the Comparative Dimension number meana standard deviation one-Sided two-Sided row total 30 32 62 non-comparative 2.6333 2.3750 2.5000 1.1290 1.2636 1.1977 34 29 63 comparative 2.6471 2.3103 2.4921 1.1250 1.0725 1.1053 Maximum (+) difference .0745 .0431 .0661 Kolmogorov-Smirnov Z .4756 .2977 .3638 I-tailed p .6361 .8376 .7674 ascores can range from O to 5, and represents the number of named brand attributes appearing in the choice criteria. 68 individuals exposed to two-sided comparative advertise- ments were not more reflective (X = 2.3103) of the named brand attributes than were the choice criteria of those individuals exposed to one-sided comparative advertisements (X = 2.6471). Hence, the findings presented in Table 13 indicate H13 Should be rejected. The research findings relating brand ownership and the incidence of named brand attributes in choice criteria are presented in Table 14. No statistically significant differences were found between the incidence of named brand attributes in the choice criteria of owners of named brands and owners of brands not named (D = .1632, Z = .7661, p = .3854), despite the fact that owners of named brands showed a slight, but consistent preference for named brand attributes as choice criteria, in both the one-sided and two-Sided cells. The findings relating to choice criteria are perhaps the most surprising; they provide no evidence to support H12, H13, or H14. In effect, they reveal that choice criteria for digital watches were not significantly altered by changes in the advertising format. In summary, Table 15 provides directionality and significance level results for all hypotheses. When empirically tested, eleven displayed ratings or measurements that were in the predicted direction; only four, however, proved to be statistically significant at the .05 level of significance. 69 TABLE 13 Choice Criteria and the Sidedness Dimension number meana standard deviation non- row comparative comparative total 30 34 64 one-Sided 2.6333 2.6471 2.6412 1.1290 1.1250 1.1183 32 29 61 two-Sided 2.3750 2.3103 2.3441 1.2636 1.0725 1.1672 Maximum (+) difference .1958 .1795 .1839 Kolmogorov-Smirnov Z .8436 .7788 1.0778 1-tailed p .2409 .2973 .0980 ascores can range from O to 5, and represents the number of named brand attributes appearing in the choice criteria. 70 TABLE 14 Choice Criteria and Brand Ownership number meana standard deviation b row one-sided two-sidedb total 12 15 27 non-owners 2.4166 2.2666 2.3333 1.1651 1.1000 1.1086 22 14 36 owners 2.7727 2.3571 2.6111 1.1104 1.0824 1.1032 Maximum (+) difference .1895 .1257 .1632 Kolmogorov-Smirnov Z .8143 .5470 .7661 1-tailed p .2786 .5922 .3854 ascores can range from O to 5, and represents the number of named brand attributes appearing in the choice criteria. bH relates only to those treatments that named competitors, thus, only the response of those subjected to comparative treatments is reported. 71 TABLE 15 Directionality and Significance Level Results Finding Hypothesis Directionality 1-tailed p INTEREST LEVEL H1 Yes .0030 H2 NO .4563 BELIEVABILITY H3 Yes .0015 H4 YeS .6528 INFORMATIONAL LEVEL H5 Yes .0004 H6 Yes .7428 SPONSOR IDENTIFICATION H7 YES .7349 H8 Yes .9895 EVOKED SET H9 Yes .0000 H10 Yes .6624 H11 Yes .9431 CHOICE CRITERIA H12 No .7674 H13 No .2973 H14 YES .3854 72 Selected Sppple Characteristics In this study, a test designed to measure pre- exposure values of the research dimensions was not utilized. It was determined that the sensitizing effects of such a test would so alert the respondents to the nature and purpose of the study that the final results would be greatly exaggerated and possibly invalid. Hence, all of the intergroup comparisons employed in this study have been based on the implicit assumption that groups were reasonably Similar (i.e., drawn from the same population) before the administration of the experimental variable. This assumption was not, however, unfounded, in part, the sample population was selected because of some known characteristics of homogeneity. The sample population was composed of students at Michigan State University taking the introductory (junior level) course in marketing, thus some reasonable conclusions about possible population similarities were drawn: most respondents would be engaged in the same primary occupation (students), would be of the same age (19-21), and would have Similar educational levels (1-3 years of college), life styles, etc. These findings were substantiated with information gained from the questionnaire (questions 15-18) which provided evidence indicating that there were no statistically significant differences, in age, sex, educa- tional achievement, or occupation, between groups. Additional measurements were taken on other, more behavioral, population characteristics. Respondents were 73 asked to answer questions on such topics as program viewership, program theme, commercial identification, program interest and information levels, brand ownership, and purchase intention. On none of these characteristics were statistically significant differences found to exist among groups. For example, over 70% of the respondents had seen "The Dinah Show" previously. Yet less than 2% had seen this particular program prior to the test. Relatedly, all groups expressed mild disagreement with the statements that this program was more informative, and more interesting, than most television programs.3 Additionally, 97% of the respondents were able to correctly identify the program theme, while 72% correctly identified the generic content of four or more commercials 4 The majority of respondents appearing in the program. owned a watch, with 38% of the owners possessing a conventional (i.e., non-digital) watch by one of the named competitors; less than 5% of all respondents expected to be buying a watch within the next six months. 3Not an altogether unexpected finding, given the forced viewing Situation, and talk-show format of the program. 4These unusually high recall scores seem directly attributable to the fact that the program theme was announc- ed a total of five times during the 30 minute program seg- ment, that a visual superimposition of the program theme (i.e., a gold record) immediately preceeded and followed each commercial break, and that immediate measurements were taken by way of (non-brand specific) aided recall devices. 74 In hindsight, then, it appears as though the sample groups were reasonably Similar before the administration of the experimental variable. Intergroup differences reported in earlier sections of this chapter, then, can properly be ascribed to the research dimensions. CHAPTER V SUMMARY AND CONCLUSIONS Introductory_Comment In this research study, hypotheses were developed linking several dimensions of communication effectiveness to a conceptual structure that classifies advertising messages according to the presence or absence of comparative and sidedness variables. An experiment was designed to itest these hypotheses, and the findings of this experimental investigation have been reported. Consistent with the framework used to report the results of the research, this chapter summarizes the research findings according to the major research dimensions of interest level, believability, informational value, Sponsor identification, composition of the evoked set, and statement of choice criteria. It concludes with recommendations for future research in related areas. Interest Level The research findings on the interest level of advertisements indicated that comparative advertisements were perceived as being significantly more interesting than 75 76 their non-comparative counterparts. Furthermore, these favorable interest level ratings for the comparative presentations were not significantly altered by systematic variation of the message Sidedness dimension. Given the consistency of the informational content in all treatments, these interest level findings seem to confirm what most advertisers have long suspected: that message structure and manner of presentation are at least as important, if not more important, than message content in creating initial awareness and interest.1 Making specific brand attribute claims, via comparison with named and recognizably presented competitors, proves to be a viable technique for increasing the interest level of a particular promotional message. Believabilipy The research findings on the believability ratings of advertisements showed that comparative advertisements were rated as being more believable than their non-comparative complements. Additionally, two-sided comparative advertisements were not viewed as being significantly more believable than one-sided comparative advertisements. The conclusion, that comparative advertisements will be viewed as being more believable than non-comparative 1It is important to note that the interest levels reported in this study are ratings of perceived interest, not physiological measures of attention or arousal. 77 advertisements, must be viewed, however, in the context of the comparative message. Comparisons were based on meaningful brand differences; differences which could be easily verified by the consumer, and which were not drastically inconsistent with existing beliefs. Nor were the compared brands disparaged. Comparative advertisements employing comparisons based on unrelated, insignificant, or irrelevant brands or differences could very well experience different results. The ability of the comparative advertisement to recreate a realistic point-of—sale situation, one that approximates real-life Shopping experiences, may be the underlying determinant Of the believability Of any comparative advertisement, not the comparative format itself. The failure Of the two-sided format to Significantly alter the believability ratings of the comparative advertisement seems to suggest that cindo? is not an important facet of believability. Two notable difficulties exist, however, which make this interpretation suspect. First, is the technical difficulty encountered in trying to incorporate a fully developed argumentative and comparative message into a 60-second time frame. Secondly, no explicit account could be taken of the extent,cgggghgy, and manner in which Opposing arguments may be present (i.e., believ- ability ratings could Suffer if two or more compared brands made irreconcilable claims for their brands). 78 Informational Value The research findings on the informational value Of advertisements revealed that comparative advertisements were consistently rated as being more informative than their non-comparative equivalents. Two-Sided comparative advertisements were not, however, rated as being Significantly more informative than one-sided comparative advertisements. Since all experimental treatments contained Similar claims for the same brand attributes, and presented these in the same order, with equal emphasis (i.e., time allotted), it can be concluded that comparative advertisements elicit more favorable informational value ratings because they present the information in a more understandable or usuable form.2 How informative any comparative advertisement is perceived, then, is most likely determined by the particular set Of brands and brand attributes employed in the compari- son; comparative advertisements that utilize the same brands and brand attributes as viewers (consumers) do, when Shopping for the product, are likely to be rated most favorably. Unfortunately, the brands and brand attributes used by the consumer vary with the urgency of the need, occasion Of product use, and product use experiences; hence, 2It is important to note that information values reported in this study are ratings Of perceived informa- tional value, not measures Of comprehension or retention Of the information presented. 79 the maximum informational effectiveness Of comparative advertisements will be achieved only when the advertiser correctly analyzes the conditions of product purchase. Sponsor Identification The research findings relating to sponsor identification disclosed that comparative advertisements did not encounter Significantly greater Sponsor misidentification than did their non-comparative counterparts. Similarly, two-Sided comparative advertisements did not encounter Significantly greater sponsor misidentification than did one-Sided comparative advertisements. These findings appear to refute the idea, held by many advertising practitioners, that comparative advertisements are not very (sales) effective because viewers may get confused and attribute the sponsor's superiority claims to one of the other named brands; however, the unusually high Sponsor identification scores are a cause Of concern. Unusually large sponsor identification scores, if artifi- cially induced, may obscure the treatment effects.3 3Sponsor identification scores may be unintentionally, yet artificially, inflated by providing for measurement of treatment effects immediately after exposure, by providing aided recall devices (e.g., multiple choice questions), and by providing an environment conducive to memorization (e.g., a university classroom); all of which are characteristic Of this study. 8O Evoked Set Research findings on the composition of the evoked sets of respondents clearly showed that the incidence of the named brands in the evoked sets of those exposed to comparative advertisements was significantly higher than the incidence Of the named brands in the evoked sets Of those exposed to non-comparative advertisements. Additionally, those exposed to two-sided comparative advertisements did not Show a significantly higher incidence Of the named brands in their evoked sets than did those exposed to one- Sided comparative advertisements. Relatedly, owners of named brands did not Show a significantly higher incidence of the Sponsored brand in their evoked sets than did the owners Of brands not mentioned. Together, these findings present a somewhat perplexing picture Of the effects of brand comparisons on the composi- tion Of the evoked set. While it was demonstrated that the use Of the comparative format increased the frequency Of identification Of named brands as relevant competitors (i.e., members of the evoked set), the infrequent occurrence of the sponsored brand in the evoked sets of those exposed to the same comparative advertisements suggests the presence Of at least one moderator variable. The named brands (Bulova and Timex), both well-known and respected brands, experienced Significantly higher evoked set designations from those exposed to the comparative treatements than from those exposed to the non-comparative treatments, while the 81 Sponsoring brand (Litronix), a relatively unknown brand, experienced infrequent evoked set designations by the same treatment groups. Hence, market stature may be a moderator variable; if so, it raises doubts about the efficacy Of making comparisons with an industry leader. Attempting to improve a brand's market stature by drawing from the prestige of better known rivals (i.e., upgrading by association) may not be effective unless this comparison established a truly Significant and demonstrable brand advantage. Choice Criteria Research findings on the statement Of choice criteria revealed no significant differences in the incidence Of named brand attributes in the choice criteria Of comparative and non-comparative viewers. Similarly, those groups exposed tO two-sided comparative messages did not evince a Significantly higher incidence of named brand attributes in their choice criteria than did those exposed to one-sided comparative messages. Relatedly, owners of named brands did not Show a significantly higher incidence Of the named brand attributes in their choice criteria than did the owners of brands not mentioned. The consistent inability Of variations in advertising format to alter the choice criteria of viewers, suggests that comparative advertisers Should restrict their brand comparisons to those brand attributes which the viewer 82 (consumer) considers most important in the selection and purchase Of the product, and should not use the comparative format to induce changes in the choice criteria; comparisons based on irrelevant brand attributes and insignificant superiority will be affected most adversely. Recommendations for Future Research While the findings Of this research study provide some revealing insights into the effectiveness Of comparative advertisements, one should resist the temptation to form sweeping generalizations from this isolated case. The results of this research study do not provide an assessment Of the effectiveness Of all types of comparative advertising, for all types Of products, in all media. Much remains to be learned about the conditions of comparative advertising effectiveness; this study offers many useful insights as to what directions future research, with sufficient time and financial resources, might proceed. Future research on comparative advertising effective— ness should proceed, for example, to investigate media influences. In particular, future inquiries should attempt to identify those media in which comparative advertising is most effective (this is an especially difficult task, Since the evaluation of the ability of an advertisement to convey the same creative strategy in different media will be largely subjective). These investigations may focus on such topics as diverse as the moderating effects Of 83 structural, time and Space, limitations, or on the role and presence of industry codes. Additionally, future research studies should attempt to expand the range of products tested and investigate and identify the relationships between product purchase characteristics, such as ShOpping effort, and comparative advertising effectiveness. Relatedly, decision rules specifying how many and which brands and attributes should be utilized to structure the comparative message could be developed and empirically tested. Future research studies could attempt to improve on the conditions of the study. Larger, more diversified, samples, for example, would enhance the probability Of gaining additional insights into the meaning and importance Of the directional findings uncovered in this study. Richer dependent measures, such as measures of arousal, attention, comprehension, retention and behavior, would upgrade the quality of the data and would permit the use of more sophisticated quantitative analysis. Likewise, viewing conditions Should represent closer approximations of real- life Situations; investigations where subjects receive multiple exposures and are subsequently confronted with contradictory and/or refutational claims would be one example. While comparative advertisements do seem to offer the advertising strategist a sound communications alternative 84 for effective communication of his message, evidence revealing the effectiveness of brand comparisons on product positioning and attainment of sales goals is still very limited. Experimental analysis Of the effects Of these conditions and factors on the performance Of comparative advertisements represents an extensive area for future work. APPENDICES .IIIIIIII. I I III \II'ZIIIIIIQ I .)I§ll'I!-n v. .. .II 1!} O. I isll '\ IONII. i APPENDIX A AUDIO/VISUAL SCRIPTS 85 86 VIDEO (1) Open on close-up of Litronix on pedestal (announcer in back- ground...out of focus) Dolly in (bringing announcer into focus) Cut to close-up Of Litronix Hold shot (as announcer demonstrates continuous second display) Cut to announcer Dolly to close-up Of announcer's face Cut to close-up Of Litronix Hold shot...superimpose "Litronix...priced below $40.00" AUDIO This is the new Litronix digital quartz crystal watch.... A product Of our space-age techno- logy...made by the people who make the insides of many of the world's electronic calculators, and backed by years Of research in quartz technology. This technological and manufacturing leadership allows you to have this find Litronix digital watch at excellent value. The Litronix digital is accurate to within ten seconds a month, and is covered by a full one-year warranty. With Litronix you don't have the nuisance Of routine maintenance associated with conventional watches There are no dials, hands, gears or springs to be damaged...but best of all, what makes the Litronix SO very special, is that the Litronix digi- tal time piece starts below $40.00! That's right...you can have this fine Litronix digital watch for less than $40.00! FIGURE A1 One-Sided Non-Comparative 87 FIGURE A2 One-Sided Comparative 83 VIDEO (1) Open on close-up of Litronix on pedestal with Bulova and Timex (announcer in back- ground...out of focus) Dolly in (bringing announcer into focus) Cut tO close-up Of Litronix Hold shot (as announcer demonstrates continuous second display) Cut to announcer Dolly to close-up of announcer's face Cut to close-up of Litronix Hold shot...superimpose "Litronix...priced below $40.00” AUDIO This is the new Litronix digital quartz crystal watch... A product of our Space-age techno- logy...made by the peOple who make the insides Of many of the world's electronic calculators. Litronix has that scientific electronics know-how that traditional watch- makers like Bulova and Timex just don't have. This technological and manufacturing leadership allows you to have this fine Litronix digital watch at excellent value. The Litronix digital is accurate to within ten seconds a month, and, like Bulova and Timex, is covered by a full one-year warranty. With Litronix you don't have the nuisance Of routine maintenance as- sociated with conventional watches. There are no dials, hands, gears, or Springs to be damaged...but best of all, what makes the Litronix SO very special is that the Litronix digital time piece starts below $40.00! That's right...you can have this fine Litronix digital watch for only a fraction of what the Bulova or Timex digitals cost! 89 VIDEO (1) Open on close-up of Litronix on pedestal (announcer in back- ground...out Of focus) Dolly in (bringing announcer into focus) Cut to close-up Of Litronix Hold shot (as announcer demonstrates continuous second diSplay) Cut to announcer Dolly to close-up Of announcer's face Cut to close-up Of Litronix Hold shot...superimpose "Litronix...priced below $40.00" AUDIO This is the new Litronix digital quartz crystal watch... A product of our space-age techno- logy...made, not by a traditional watchmaker, but by the peopTe who make the insides Of many of the world's electronic calculators. This technological and manufacturing leadership allows you to have this fine Litronix digital watch at excellent value. The Litronix digital is accurate to within ten seconds a month, and is covered by a full one-year warranty. Of course batteries must be replaced occasionally, but, with the Litronix you don't have the nuisance of routine maintenance associated with conventional watches. There are no dials, hands, gears or springs to be damaged...but best Of all, what makes the Litronix so very special, is that the Litronix digital time piece starts below $40.00! Inexpensive, but not cheap...no frills, just solid split-second accuracy, at a price you can afford! FIGURE A3 Two-Sided Non-Comparative 90 FIGURE A4 Two-Sided Comparative 91 VIDEO (1) (7) Open on close-up Of Litronix on pedestal with Bulova and Timex (announcer in back- ground...out of focus) Dolly in (bringing announcer into focus) Cut to close-up Of Litronix Hold shot (as announcer demonstrates continuous second hand display) Cut to announcer Dolly to close-up of announcer's face Cut to close-up of Litronix Hold shot...superimpose "Litronix...priced below $40.00" AUDIO This is the new Litronix digital quartz crystal watch... A product of our space-age techno- logy...made by the peOple who make the insides Of many Of the world's electronic calculators. Litronix has that scientific electronics. know-how that OOd traditional watchmakers like Bu ova and Timex just don't.have. This technological and manufacturing leadership allows you to have this fine Litronix digital watch at excellent value. The Litronix digital is accurate to within ten seconds a month, and, like Bulova and Timex, is covered by a full one-year warranty. Of course batteries must be replaced occasionally...but with the Litronix you don't have the nuisance Of routine maintenance associated with conventional watches... There are nO dials, hands, gears or springs to be damaged...but best Of all, what makes the Litronix SO very special, is that the Litronix digital time piece starts below $40.00! Inexpensive, but not cheap..no frills, just solid split-second accuracy, for only a fraction of what the Bulova or Timex digitals cost! APPENDIX 8 PROGRAM CHARACTER AND SEQUENCE 92 93 8 Minute Show Segment 6O 3O 11 6O 6O 30 .Dinah sings and introduces guests Second Commercial Spot .Youth Basketball Association Second Commercial Spot .Community Calendar Minute Show Segment .Marty Robbins Sings and is interviewed by Dinah Second Commercial Spot .Litronix Digital Watch Second Commercial Spot .National Association of School Administrators Second Commercial Spot .Climalene Laundry Detergent Additive 6 Minute Show Segment Temptations singing FIGURE 81 Program Character and Sequence APPENDIX C QUESTIONNAIRE 94 95 TELEVISION COMMUNICATIONS STUDY You have just seen a portion Of the Dinah Show, recorded at an earlier date. I would like you to answer a few questions relating to this Show and the advertisements presented in the Show. Individual participation in this study is anonymous, therefore you do pg; have to identify yourself by Signing the questionnaire. Program Date Time 96 Please read each of the following questions or statements and circle the response that best expresses your feeling about the question or statement. Please do not mark in the column on the right hand side of the paper. 1. Have you seen the Dinah Show before? (5-6) 1. yes 2. no If ygg, have you seen this particular program before? (7-8) 1. yes 2. no 2. The theme for the program was? (9-10) 1 Dinah Salutes Country Music 2. Dinah and the Gold Record Winners 3. Dinah and Singers of the 60's 4 Dinah and Grammy Award Winners , 5 Other (please specify) 3. Please identify those advertisements that you (11-16) believe appeared in the program. . An advertisement for a laundry detergent additive. An advertisement for a body deodorant. An advertisement for a youth basketball association. An advertisement for a cat food. An advertisement for a digital watch. An advertisement for school administration/education. An advertisement for orange juice. An advertisement for the national guard. CDVONU‘I-wao—I O C O O Q C 0 With the following scheme in mind, please respond to questions 4-8 by circling that number that best expresses your feeling about that question or statement. . Strongly agree agree agree Sqnewhat neither agree, nor disagree disagree somewhat disagree strongly disagree \IOWU‘l-bLUNH 4. The program was more entertaining than most television (17-18) programs. strongly ‘ strongly agree disagree 1 2 3 4 5 6 97 The program was more informative than most television (19-20) programs. strongly strongly agree disagree _1.__ _2_ _3__ L _5_ _6_ _7_. The advertisement for the digital watch was more (21-22) interesting than most television advertisements. strongly strongly agree disagree _1_ _2_ _§_ _4_ _5_ _5_ __ The advertisement for the digital watch was more (23-24)_ informative than most television advertisements. strongly strongly agree disagree L L _3__ .3. _5_ _§_._7_ The advertisement for the digital watch was more (25-26) believable than most television advertisements. strongly strongly agree disagree .1. __2_ _3__ _4_ .2._6__7_ In the advertisement for the digital watch, what (27-30) points were stressed by the sponsor? 1. water resistant shock resistant 3 anti-magnetic 4 nO-glare crystal 5. time display system 6. functions performed 7. type Of power source 8 ease of maintenance 9. repair/service facilities 10. accuracy 11. styling 12. guarantee/warranty 13. reputation for watchmaking excellence 14. reputation for electronics excellence 15. availability 16. low price 17. other (please Specify) 98 The sponsor of the advertisement for the digital (31-32) watch was: 1. Benrus 13. Hamilton 2. Longines Wittnauer 14. Speidel 3. Fairchild 15. Omega 4. Ultimatic 16. Litronix 5. Bulova 17. Tissot 6. Westclox 18. Microsonic 7. Concord 19. Timex 8. Pulsar 20. Elgin 9. Novus 21. Texas Instruments 10. Waltham 22. Rolex 11. Seiko 23. Other (please specify) 12. Baylor Please identify three major competitors in the digital (33-34) watch market. 1. Rolex 13. Novus 2. Elgin 14. Seiko 3. Microsonic 15. Benrus 4. Litronix 16. Texas Instruments 5. Speidel 17. Concord 6. Baylor 18. Bulova 7. Waltham 19. Hamilton 8. Pulsar 20. Omega 9. Westclox 21. Tissot 10. Timex 22. Ultimatic 11. Longines Wittnauer 23. Other (please specify) 12. Fairchild What product characteristics do you consider to be (35-38) most important in the selection and purchase Of a digital watch? 1. availability 2 reputation for watchmaking excellence 3. guarantee/warranty 4. accuracy 5 6 7 8 ease of maintenance functions performed no-glare crystal . Shock resistant 9. low price 10. reputation for electronics excellence 11. styling 12. repair/service facilities 13. type Of power source 14. time display system 15. anti-magnetic 16. water resistant 17. Other (please specify) 99 13. DO you presently own a watch? (39-40) 1. yes 2. no If as, what brand is it? (41—42) If 123; is it a digital watch? (43-44) 1. yes 2. no 14. DO you anticipate buying a digital watch within the (45-46) next Six months? 1. yes 2. no If ygg, will you be buying the watch for: (47-48) 1. yourself 2. someone else If ygg, what brand would you purchase? (49-50) Finally, I would like to Obtain some personal information about you. 15. Please circle that number that corresponds to your (51-52) age. 1. under 24 years 2. 25-34 years 3. 35-44 years 4. 45-54 years 5. 55-64 years 6. 65 years and Older 16. Please circle that number that corresponds to the (53-54) highest grade level you have achieved. Less than twelfth grade 2 High school graduate 3. 1-3 years of college 4 5 t—l O 0 College graduate Some graduate training 17. Please circle that number that corresponds to your (55-56) sex. 1. male 2. female 100 18. Please circle that number that corresponds to your (57-58) occupation. 1. Craftsman 2. Professional, manager or administrator 3. Service worker 4. Clerical worker 5. Sales worker 6. Laborer 7. Housewife 8. Student 9. Retired 10. Unemployed I would like to take this opportunity to thank you for your cooperation and participation in this study. BIBLIOGRAPHY This bibliography consists of three sections: Books, Monographs, and Dissertations; Academic Journals and Conference Proceedings; and, Periodicals and Newspapers. The first two sections are organized alphabetically according to the surnames Of the authors. The latter is presented with subheadings for each publication, under which the references follow in chronological order. BIBLIOGRAPHY BOOKS, MONOGRAPHSg AND DISSERTATIONS Abramson, Joseph. "Comparative Advertising, Inoculation Theory, and the Prevention of Attitude Change Among Brand Loyal Consumers: A Laboratory Experiment," (Ph.D. Dissertation, The Louisiana State University and Agricultural and Mechanical College, 1977). Baird, Thomas R. "An Experimental Study Of the Effects Of Information From Advertisements and Consumer Rating Publications on Criteria Listed by Consumers as Important to Product Selection," (Ph.D. Dissertation, University Of Arkansas, 1977). Campbell, Brian Milton. "The Existence of Evoked Set and Determinants of its Magnitude in Brand Choice Behaviorfl' in Buyer Behavior: Theoretical and Empirical FoundatiOns, JohnTA. Howard and Lyman E.‘OstTund, Eds., (New York: Alfred A. Knopf, 1973), 243-244. Chaffee, Steven H., and Jack M. McLeod. "Consumer Decisions and Information Use," in Consumer Behavior: Theoretical Sources, (EnglewOOdCliffS, N.J.: Prentice- Hall, 1973), 385-415. Engel, James F., David T. Kollat, and Roger D. Blackwell. Consumer Behavior, (Hinsdale, Illinois: Dryden Press, 2nd. Edition, 1973). Golden, Linda L. "Consumer Reactions to Direct Brand Comparisons in Advertisements," (Ph.D. Dissertation, University Of Florida, 1975). Hovland, Carl 1., Arthur A. Lumsdaine, and Fred A. Sheffield. Experiments on Mass Communication, (Princeton, N.J.: Princeton University Press, 1949). Howard, John A., and Jagdish N. Sheth. The Theory Of Buyer Behavior, (New York: John Wiley & Sons, Inc., 1969T. Howard, John A., and James Hulbert. Advertising and the Public Interest, (Chicago: Crain Communications Inc., 1973). 101 102 Kerlinger, Fred N. Foundations of Behavioral Research, (New York: Holt, Rinehart and Winston, Inc., 1973). Koehler, J. W. "Effects on Audience Opinion of One-Sided and Two-Sided Speeches Supporting and Opposing a Proposition," in The Process Of Social Influence, Thomas D. Beisecker and Donn W. Parson, Eds., (Enggewood Cliffs, N.J.: Prentice-Hall, 1972), 351- 69. Policy Statement and Guidelines for Comparative Advertising, (New York: American Association of Advertising Agencies, Bulletin Number 3151, April 3, 1974). Sellars, Ronald Kay. "A Study Of the Effectiveness Of Comparative Advertising for Selected Household Appliances," (Ph.D. Dissertation, The Louisiana State University and Agricultural and Mechanical College, 1977 . Siegel, Sidney. Nonparametric Statistics, (New York: McGraw-Hill Book Company, 1956). Spence, Homer E., and James F. Engel. "The Impact of Brand Preference on the Perception of Brand Names: A Laboratory Analysis," in Research in Consumer Behavior, David T. Kollat, Roger D. Blackwell, and James Engel, Eds., (New York: Holt, Rinehart & Winston, 1970), 61-70. Ulanoff, Stanley M. Comparative Advertising: An Historical Retrospective, (Working Paper), (Cambrid e, Mass: Marketing Science Institute, February 1975?. U. S. Senate, Subcommittee on Executive Reorganization and Government Research on the Committee on Government Operations. Hearings on the Federal Role in Consumer Affairs, 9lst. Cong., 2nd. Sess., (1970). ACADEMIC JOURNALS AND CONFERENCE PROCEEDINGS Barry, Thomas E., and Roger L. Tremblay. "Comparative Advertising: Perspectives and Issues," Journal of Advertising Research, (November 1975), 15-20. Bither, Stewart W., Ira J. Dolich, and Elaine B. Nell. "The Application of Attitude Immunization TechniQueS in Marketing," Journal of Marketing Research, (February 1971), 56-61. 103 Boyd, Harper W., Michael L. Ray, and Edward C. Strong. "An Attitudinal Framework for Advertising Strategy," Journal of Marketing, (April 1972), 27-33. Brown, Steven W., and Donald W. Jackson. "Comparative Television Advertising: Examining its Nature and Frequency," Journal of Advertising, (November 1977), 15-18. Bucklin, Louis P. "The Informative Role of Advertising," Journal Of Advertising Research, (September 1965), 11-15. Chevins, Anthony C. "A Case for Comparative Advertising," Journal of Advertising, (Spring 1975), 31-36. Chu, Godwin C. "Prior Familiarity, Perceived Bias, and One-Sided Versus Two-Sided Communications," Journal of Experimental Social Psychology, (July 1967), 243-254. Clancy, Kevin J., and Lyman E. Ostlund. "Commercial Effectiveness Measures," Journal Of Advertising Research, (February 1976), 29-34. Faison, Edmund Winston Jordan. "Effectiveness of One-Sided and Two-Sided Mass Communications in Advertising," Public Opinion Quarterly, (Fall 1961), 468-469. Greenberg, Allan. "IS Communications Research Really Worthwhile?" Journal of Marketing, (January 1967), 48-50. Greenland, Leo. "Advertisers Must StOp Conning Consumers," Harvard Business Review, (July-August 1974), 18-20. Hansen, Robert A., and Carol A. Scott. "Comments on 'Attribution Theory and Advertiser Credibility'," Journal Of Marketing Research, (May 1976), 193-197. Hass, Glen R., and Darwyn E. Linder. "Counterargument Availability and the Effects of Message Structure on Persuasion," Journal of Personality and Social Psychology, (Vol. 23, 1972), 219-233. Hulbert, James. "Information Processing Capacity and Attitude Measurement," Journal of Marketing Research, (February 1975), 104-106. Insko, Chester A. "One-Sided Versus Two-Sided Communications and Countercommunications," Journal Of Abnormal and Social Psychology, (September 1962), 203-206. 104 Jacoby, Jacob, Donald E. Speller, and Carol Kohn Berning. "Brand Choice Behavior as a Function Of Information Load: Replication and Extension," Journal Of Consumer Research, (June 1974), 33-42. Jacoby, Jacob,Donald E. Speller, and Carol A. Kohn. "Brand Choice Behavior as a Function of Information Load," Journal Of Marketing Research, (February 1974), 63-69. Jones, Russell A., and Jack W. Brehm. "Persuasiveness Of One- and Two-Sided Communications as a Function of Awareness there are Two Sides," Journal of Experimental Social Psychology, (January 1970), 47-56. Kassarjian, Harold H., and Joel B. Cohen. "Cognitive Dissonance and Consumer Behavior," California Management Review, (Fall 1965), 55-64. Lamb, Charles W., Jr., William M. Pride, and Barbara A. Pletcher. "A Taxonomy for Comparative Advertising Research," Journal of Advertising, (February 1978), 43-47. Levine, Philip. "Commercials That Name Competing Brands," Journal Of Advertising Research, (December 1976), 7-14. Lumsdaine, Arthur A., and Irving L. Janis. "Resistance to 'Counterpropaganda' Produced by One-Sided and Two- Sided 'Propaganda' Presentations," Public Opinion Quarterly, (Fall 1953), 311-318. Maloney, John C. "Is Advertising Believability Really Important?" Journal Of Marketing, (October 1963), 1-8. Mazis, Michael B. "A Framework for Evaluating the Effects Of Comparative Advertising," (Paper presented to the American Marketing Association Educators' Conference, Rochester, N.Y., August 1975). Oshikawa, Sadaomi. "Can Cognitive Dissonance Theory Explain Consumer Behavior?" Journal Of Marketing, (October 1969), 44-49. Prasad, V. Kanti. "Communications - Effectiveness of Comparative Advertising: A Laboratory Analysis," Journal Of Marketing Research, (May 1976), 128-137. Roering, Kenneth J., and Robert J. Paul. "The Effect Of the Consistency of Product Claims on the Credibility of Persuasive Messages," Journal Of Advertising, (Spring 1976), 32-36. 105 Sawyer, Alan G. "The Effects of Repetition Of Refutational and Supportive Advertising Appeals," Journal of Marketing Research, (February 1973), 23-33. Sears, David 0., and Jonathan L. Freedman. "Selective Exposure to Information: A Critical Review," Public Opinion Quarterly, (Summer 1967), 194-213. Settle, Robert B., and Linda L. Golden. "Attribution Theory and Advertiser Credibility," Journal of Marketing Research, (May 1974), 181-185. Shimp, Terence A. "Comparison Advertising in National Television Commercials: A Content Analysis," (Paper presented to the American Marketing Association Educators' Conference, Rochester, N.Y., August 1975). Steiner. "Receptivity to Supportive Versus Nonsupportive Communications," Journal Of Abnormal and Social Psychology, (October 1962), 266-267. Summers, John 0. "Less Information is Better?" Journal of Marketing Research, (November 1974), 467-468. Szybillo, George J., and Richard Heslin. "Resistance to Persuasion: Inoculation Theory in a Marketing Context," Journal of Marketing Research, (November 1973), 396-403. Tannenbaum, Percy H. "The Indexing Process in Communication," Public Opinion Qnarterly, (Summer 1955), 292-302. Wilkie, William L. "Analysis of Effects of Information Load," Journal Of Marketing Research, (November 1974), 462—466. Wilkie, William L., and Paul W. Farris. "Comparison Advertising: Problems and Potential," Journal of Marketing, (October 1975), 7-15. Wilkie, William L., and David M. Gardner. "The Role Of Marketing Research in Public Policy Decision Making," Journal of Marketing, (January 1974), 38-47. Williams, Terrell G. "An Experiment in the Effects of Comparative Advertising on Brand Awareness, Comprehen- sion, and Preference," Southern Business Review, (Spring 1978), 30-42. 106 Winter, Frederick W. "A Laboratory Experiment of Individual Attitude Response to Advertising Exposure," Journal Of Marketing Research, (May 1973), 130-140. Wright, Peter L. "The Cognitive Processes Mediating Acceptance of Advertising," Journal Of Marketing Research, (February 1973), 53-62. PERIODICALS AND NEWSPAPERS Advertising Age Tyler, William 0. "Is Competitive Comparison Really Bad in Advertising? Reform With Care," Advertising Age, (March 14, 1966), p. 61. "Naming Competitors in Ads," (Editorial), Advertising Age, (March 21, 1966), p. 20. Diamond, Sidney A. Reference to Competitive Product by Name Can be Trademark InfringementJ'Advertising Age, (Febru- ary 6, 1967), 81-82. Haugh, Louis J. "TWA Again Revises On-Time Ad, Starts Tackling Overseas Lines," Advertising_Age, (March 22, 1967), 2+. Diamond, Sidney A. "Court Okays Use Of Competitor's Trademark in Ads," Advertising Age, (December 16, 1968), 66. Diamond, Sidney A. "Chanel Wins Only Partial Victory in Unfair Competition Case," Advertising Age, (November 5, 1973). p. 54. Grant, Don. "Be Hard on Comparison Ads: Roberts to Four A's," Advertising Age, (November 19, 1973), 1+. "NARB Rules Against Schick Shaver Ads, Cites 'Implications'," AdvertisingyAge, (December 31, 1973), p. 1+. "Competitors Hail NARB for Schick Shaver Ruling," Adverti- sing Age, (January 7, 1974), p. 1+. "Comparative Ads Valid - If," (Editorial), Advertising Age, (January 21, 1974), 16. Christopher, Maurine. "NBC Spells Out New Formal Guides for Comparative Spots," Advertising Age, (January 28, 1974), 1+. 107 "Advertisers' Name-Calling Could Hurt, McMahan Says," Advertising Age, (February 18, 1974), 72. "TOO Much Info in Ads Hurts Consumer, Advertiser: Wells," Advertising Age, (March 4, 1974), 57. Christopher, Maurine. "ABC Comparative Rules Stress Test Procedures," Advertising Age, (March 18, 1974), 1+. "NARB Rules Carte Blanche Comparison Ads Misleading," Advertising Age, (July 22, 1974), 16. "Care in Comparisons Urged; Cosmetics Market Tacks Hit," AdvertisingyAge, (August 5, 1974), 31-32. "FTC Priority is Getting Full Info, Facts in Ads, Engman Tells Lawyers," Advertising Age, (August 19, 1974), 4. Giges, Nancy. "P & G's Newest Ad Efforts Indicate Move to More Competitive Approach," Advertising Age, (September 16, 1974), 1+. "Review Board Sets Guides on Comparisons," Advertising Age, (October 7, 1974), 8. "NBC Sets Guidelines for Comparative Ads Plaints," Adverti- sing Age, (January 6, 1975), 8. "Earth Born Comparative Ads Halted, NAD Reports," Advertising Age, (Janaury 20, 1975). Giges, Nancy. "Warner-Lambert Presses NAB on Scope 'Medi- cine Breath' Ads," Advertising Age, (February 24, 1975), 1+. "Underdog Advertiser Wins in 'Naming Names': 8800," Advertising Age, (March 10, 1975), 56. "ABC Censor Raps Trend to Naming Rivals in Ads," Advertising Age, (March 31, 1975), 1+. Tyler, William 0. "Comparative Advertising: A Powerful Selling Tool When it is not Abused," Advertising Age, (April 21, 1975), 58-60. "Open Way for Comparative Ads, Canada Meet Told," Adverti- sing Age, (May 12, 1975), 6. "Toyota Quality Compared to Other Imports in Summer Ads," Advertising Age, (June 9, 1975), 2. Christopher, Maurine. "NBC Widens Comparative Price Ad Rules," Advertising Age, (September 1, 1975), 42. 108 "Comparative Ads Ineffective: 0 & M Study," Advertising Age, (October 13, 1975), 16. "Comparative Ads a 'Disaster'?" (Editorial), Advertising Age, (October 20, 1975), 14. Donath, Bob. "Comparative Ad Study Sparks Debate," Advertising Age, (November 3, 1975), 3+. "NARB Toughens Comparative Ad Rules," Advertising Age, (December 1, 1975), 1+. Cohen, Stanley. "Widespread FTC Probe Will Seek Codes That Hinder Comparative Ads," Advertising Age, (February 23, 1976), 1+. "Comparative Ads on TV Rise, Researcher Says," Advertising Age, (March 1, 1976), 67. Haugh, Louis J. "TWA Widens Comparative Ad Use Despite More Industry, CAB Pressure," Advertising Age, (March 15, 1976 , 2. Bernstein, Sid. "Comparison Ads Get Out of Hand," Advertising Age, (March 29, 1976), 16. “Those Competitive Airlines." (Editorial), Advertising Age, (March 29, 1976), 16. "Tylenol Exec Speaks Out on Datril Price Ad," Advertising Age, (March 29, 1976), 1+. "Coca-Cola, Pepsi Put Taste Tests Into Comparison Arena," Advertising Age, (April 5, 1976), 1+. "Tannenbaum: Comparative Ads Can Work; Kershaw Says NO," Advertising Age, (May 17, 1976), 1+. "Comparative Ads Still Beckon," (Editorial), Advertising Age, (May 24, 1976), 14. Giges, Nancy. "Coke Ad Hits Pepsi Comparison, Hints Blind Taste Test Not Valid," Advertising Age, (June 14, 1976), p0 1+. Cohen, Stanley E. "FTC Says Guides Near on Ads That Name Names," Advertising Age, (June 28, 1976), p. 3+. "Comparison Ads Will Get Little FTC Interference," Advertising Age, (July 5, 1976), 8+. 109 Kershaw, Andrew G. "For and Against Comparative Advertising - For," Advertising Age, (July 5, 1976), 25-26+. Tannenbaum, Stanley I. "For and Against Comparative Advertising - For," Advertising Age, (July 5, 1976), 25-26+. "World Book Takes on the Big Guy, Article by Article," Advertising Age, (July 12, 1976), 4. ”Comparison Ads Bubble in Canada," Advertising Age, (August 9, 1976), 2+. Haugh, L. J. "TWA Directing Its Comparative Story to Business Travelers," Advertising Age, (September 20, 1976), 3+. Chew, Robert. "Coke's Halt of Counter-effort Brings Cheer," Advertising Age, (November 15, 1976), 1+. Giges, Nancy. "Competitor Ads Take Off After Tylenol," Advertising Age, (December 6, 1976), 1+. Giges, Nancy. "Coca-Cola - Reluctant Entrant Into Comparative Ad Warfare," Advertising Age, (January 3, 1977), 2+. O'Connor, John J. "Cigar Sales Slide Drives Marketers to Competitive Thrust," Advertising Age, (February 21, 1977), 2+. "NAD Has the Final Word on Soft Drink Battle of '76," Advertising Age, (May 16, 1977), 8. "Coke Renews Attack on Pepsi Taste Test," Advertising Age, (May 23, 1977), 1+. Krakowiecki, Marie. "What Else is Going Up? Ad Liability Insurance Cost," Advertising Age, (August 15, 1977), 32+. "Bickering, Fear Of FTC Rendered Report Toothless," Advertising Age, (August 22, 1977), 62. Giges, Nancy. "Judge Rules for Tylenol; Anacin Ordered to Halt Inflammation Ad Claims," Advertising Age, (August 22, 1977), 1+. Giges, Nancy. "NARB Paper Ducks Conclusion on Comparative Ads," Advertising Age, (August 22, 1977), 62. 110 "Comparative Ads Grade 'F' by CBBB President," Advertising Age, (October 17, 1977), 2+. "Comparative Advertising Attracts Divergent Views," Advertising Age, (November 21, 1977), 84. Keddy, Barbara. “Comparative Ads Still a Challenge," Advertising Age, (February 13, 1978), p. 86. "Coke Calorie Claim Clouts Challenger," Advertising Age, (April 24, 1978), 2+. Broadcasting "NBC Puts on Paper Its Rules for Comparative Ads," Broadcasting, (February 4, 1974), 40-41. "Code Review Board Adds More Teeth to TV Rules for Comparison Ads," Broadcasting, (October 7, 1974), 61. "Comparative Ads in Center Ring at AAAA Meeting," Broadcasting, (May 17, 1976), 42-44. "Study Finds That Naming Names in TV Ads Isn't Always Beneficial," Broadcasting, (February 14, 1977), 76-77. "Study Cites Value Of Comparative Ads But Warns Effect Hinges on Honesty," Broadcasting, (August 29, 1977), 52. Reader, Peter H. "Pepsi Takes Coke Head-on in Television Taste Tests," Broadcasting, (February 20, 1978), 14. Editor & Publisher ”Effects of Comparative Ads Studied," Editor & Publisher, (January 21, 1978), 22. Fortune Morner, Aimee L. "It Pays to Knock Your Competitor," Fortune, (February 13, 1978), 104-106+. Industrial Marketing Sawyer, Howard G. "800 to the 'One Book Buy,‘ Ads for Competition, and 'Right'?" Industrial Marketing, (November 1976), 55. 111 Marketing News Woodside, Arch G. "Comparative Ads Help Consumers if Test Results are Told, But Ultra Ban 5000 Ad Claims Fail Student Test," Marketing News, (November 15, 1974), 5. New York Times Dougherty, Philip H. "Calling Brand X by its Real Name," New York Times, (January 21, 1973), 15. Dougherty, Philip H. "Advertising: Comparison Issue," New York Times, (November 14, 1973), 61. Printers' Ink "Naming Competitors in Ads: Forthright, Fair, Foolish?" (Commentary). Printers' Ink, (Janaury 28, 1966), 32-34. Sales & Marketinnganagement Trytten, John M. "It's as Easy as Pie: Nothing Can Compare With a Bad Comparative Ad," Sales & Marketing Management, (July 12, 1976), 61-62. Television and Radio Age "Four A's Does an About-Face in Comparative Ad Guidelines," Television/Radio Age, (April 15, 1974), 49. Bradshaw, Tom. "Comparative Ads: What's Their Status Now?" Television/Radio Age, (April 29, 1974), 29-32+. Time "Naming Names," Time, (August 20, 1965), 70-71. The Wall Street Journal "SO Long, Brand X: Naming Names of Rival in Ads is Catching on but Spurs Controversy," The Wall Street Journal, (December 26, 1973), 1.